A comparison of the Fishbein-Ajzen and Bentler

University of Nebraska at Omaha
DigitalCommons@UNO
Student Work
8-1981
Attitude-behavior relationships: A comparison of
the Fishbein-Ajzen and Bentler-Speckart models
Arlene J. Fredericks
University of Nebraska at Omaha
Follow this and additional works at: http://digitalcommons.unomaha.edu/studentwork
Part of the Psychology Commons
Recommended Citation
Fredericks, Arlene J., "Attitude-behavior relationships: A comparison of the Fishbein-Ajzen and Bentler-Speckart models" (1981).
Student Work. Paper 92.
This Thesis is brought to you for free and open access by
DigitalCommons@UNO. It has been accepted for inclusion in Student
Work by an authorized administrator of DigitalCommons@UNO. For
more information, please contact [email protected].
A T T I T U D E - B E H A V I O R RELATIONSHIPS:
A C O M P ARISON OF THE FISHBEIN-AJZEN
A N D BENTLE R - S P E C K A R T MOD E L S
A Thesis
Presented to the
De p artment of Psychology
and the
F a c u l t y of the Graduate College
U n i v e r s i t y of N e b raska
In Partial Fulfillment
of the R e q u irements for the Degree
M a s t e r of Arts
U n i v e r s i t y o f N e b r a s k a at Om a h a
by
Arl e n e J. Fredricks
August
1981
UMI Number: EP72742
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
UMI EP72742
Published by ProQuest LLC (2015). Copyright in the Dissertation held by the Author.
Microform Edition © ProQuest LLC.
All rights reserved. This work is protected against
unauthorized copying under Title 17, United States Code
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
THESIS ACCEPTANCE
A c c e p t e d for the faculty of the Graduate College,
Nebraska,
University of
in p a r tial f u l f illment of the requirements for the degree
M a s t e r of Arts,
U n i v e r s i t y of N e b r a s k a at Omaha.
Thesis Committee
ame
Department
Chairman
Date
ii
ACKNOWLEDGEMENTS
The author is indebted to D. Kenneth Spenner of the Boys Town
C e n t e r for Y o u t h D e v e l o p m e n t for not only m a k i n g the LISREL computer
p r o g r a m available, but for his invaluable instruction and guidance that
m a d e use of it possible;
guidance,
Dr. J.
Dr. Dennis Dossett for innumerable suggestions,
editorial assistance,
Brad Chapman,
and helpful comments;
and encouragement;
Dr. Carl Greenberg,
Mr.
Dr.
the C o m m i t t e e - -
Gaylon Oswalt,
P a trick Jor d a n for coding data;
for reading
Instructors
who c ollected d aily a t tendance data and per m i t t e d the use of class time
for ques t i o n n a i r e a d m i n i s t r a t i o n - - M r . David Arnold,
Mr.
W i l l i a m Grisham,
R a y m o n d Millimet,
Mr.
G a r y Kinstlinger,
and Mr. E d w a r d Ward;
Dr.
Barbara Manning,
J o s e p h LaVoie, Mr. J a c k Leon, Mr.
Oswalt;
Dr. C.
Instructors who p e r m i t t e d use of
class time for p ilot sample da t a collection--Dr.
Dr.
Dr. Carl Greenberg,
Kenneth D e f f e n b a c h e r ,
Russell Montgomery,
Dr.
Gaylon
and Mr. J i m Smith of the UNO Computer Netw o r k for m u c h a p p r e ­
ciated technical assistance.
TABLE OF CONTENTS
Page
T.i st. of T a b l e s . ...........................................................
vi
List of F i g u r e s .............................................................. vii
A b s t r a c t .................................................................... viii
The A t t i t u d e - B e h a v i o r R e l a t i o n s h i p .......................................
1
The Fishbein-Aj zen M o d e l ..........................................
3
Conceptual F r a m e w o r k of the Fishbein-Aj zen M o d e l ..............
5
The Fishbein-Aj zen Model of At t i t u d e F o r m a t i o n ................
7
Fishb e i n and A j z e n ’s Specifications for M e a surement
9
.........
The B e n tler and Speckart M o d e l ...........................................
12
M e t h o d .......................................................................
17
S u b j e c t s .....................
17
P r o c e d u r e ...........................................................
18
M e a s u r e s .............................................................
20
A n a l y s i s .............................................................
22
R e s u l t s .....................................................................
24
Pilot and Prim a r y Study Sample C o m p a r i s o n s .....................
24
M e a s u r e m e n t M o d e l ..................................................
24
Structural M o d e l ....................................................
25
Model C o m p a r i s o n s ..................................................
25
C h i - S q u a r e Good n e s s of Fit T e s t s ................
34
C h i - S q u a r e D i f f e r e n c e Tests
......................................
37
D i s c u s s i o n .............. ....................................................
40
iv
Page
R e f e r e n c e N o t e ................
46
R e f e r e n c e s ..................................................................
47
Appendix A--Questionnaire
for Pilot Study
50
Appendix B--Questionnaire
for Primary Study
.............................
...........................
55
A p p e n d i x C - - T a b l e s of Results ............................................
69
A p p e n d i x D - - D i s c u s s i o n of
82
Path Analysis
v
..............................
LIST OF TABLES
T a ble
Page
1*
Structural
M odel S p e c i f i c a t i o n s .................................... 29
2.
Path C o m parisons b e t w e e n Models, t-Values,
3.
Chi-S q u a r e G o o d n e s s - o f - F i t Tests
4.
Chi-S q u a r e D i f f e r e n c e Tests for Model C o m p a r i s o n s ................ 38
vi
. . .
.............
.....................
31
35
LIST OF FIGURES
Figure
1.
2.
3.
4.
Page
S chematic R e p r e s e n t a t i o n of Fishbein - A j z e n
M odel of A t t i t u d e - B ehavior R e l a t i o n s h i p .......................
Schematic R e p r e s e n t a t i o n of Fish b e i n - A j z e n Model
and of the Bent l e r - Speckart M o d i f i c a t i o n . . .
6
...............14
F i s h b e i n - A j z e n M o del (Prior Behavior Excluded)
w i t h Paths Labeled for LISREL An a l y s i s .......................
26
B e n t l e r - S p e c k a r t Model (Prior Behavior Included)
with Paths Labeled for LISREL Analysis .......................
27
vii
AB S T R A C T
This study compared the F i s h bein - A j z e n
r e l a t i onships with B e n t l e r - S p e c k a r t 's
(1975) model of attitude-behavior
(1979) m o d ifications of the model.
Subjects were 236 u n d e r g r a d u a t e college students and the measures of
b e h a v i o r were r e p e a t e d s elf-reports of class attendance.
linear structural relationships,
lying construct,
usi n g m u l tiple
An analysis of
indicators for each u n d e r ­
supported the Bentler-Speckart addition to the Fishbein-
Aj zen model of p r i o r b e h a v i o r as a direct causal
sequent b e h a v i o r and behav i o ral
original F i s h b e i n - A j z e n model,
sequent b e h avioral
intentions.
influence on both sub­
However,
consistent w i t h the
a direct causal p a t h from attitude to s u b ­
intentions was not found.
Directions for future studies
and r e s p e c i f i c a t i o n of the model were discussed.
viii
1
T h e A t t i t u d e - B e h a v i o r Relationship
That a ttitude can sometimes pred i c t behavior has be e n documented
and r e po r t e d in recent reviews of the a t t i t u d e-behavior relationship
literature
(Ajzen § Fishbein,
Himmelfarb,
1978;
Kelman,
1977; Calder § Ross,
1974).
However,
1977;
Eagly £
d i s c overy of the conditions
a nd p r oc e s s e s that perm i t p r e d ictions of be h a v i o r remains a challenge
for behav i o r a l scientists.
This r e s earch quest i o n must neces s a r i l y be
p a r t i t i o n e d into a c o n c e p t u a l i z a t i o n of attitudes on the one hand and
b e h a v i o r on the other.
measured,
Since be h a v i o r is mo r e easily obser v e d and
it is the a ttitude construct that has att r a c t e d the greatest
a m ount of m e t h o d o l o g i c a l attention.
N u m e r o u s defin i t i o n s of attitude h a v e b e e n promulgated.
However,
the p r e s e n t d i s c u s s i o n will be limited to only rela t i v e l y recent
approaches.
For example,
R o k e a c h defined attitude as "an organization
of inter r e l a t e d b e l iefs around a common object, w i t h certain aspects of
the object b e i n g at the focus of attention"
Tr i a n d i s
(1968, p.
116).
(1971) p r e s e n t e d a definition wh i c h he felt included m a n y
p r e v i o u s l y d e v e l o p e d c e n tral ideas as follows:
"An a t t itude is an idea
char g e d w i t h emot i o n w h i c h p r e d i s p o s e s a class of actions to a p a r t i c u ­
lar class of social s i t u a t i o n s "
three components of attitude:
(p. 2).
This d e f i n i t i o n references
(a) the cognitive or "idea" component;
(b) the a f f e c t i v e or emotional component;
p r e d i s p o s i t i o n to action component.
and
(c) the behavioral or
In this context p r e d i s p o s i t i o n to
a cti o n does not n e c e s s a r i l y imply actual behavior.
in d i s c u s s i n g the a t t i t u d e - b e h a v i o r relationship,
involve:
T r i andis
(1971),
notes that attitudes
2
W hat p e o p l e think about,
feel
about,
and how they would like
to b e have toward an attitude object.
Behavior is not only
d e t e r m i n e d b y w h a t p e ople would like to do but also by what
they think they should do,
h a v e u s u a l l y done,
that is, social n o r m s , b y what they
that is, h a b i t s , and b y the expected
co n s e quences of the b e h a v i o r .
(p.
14)
Implicit in this d e f i n i t i o n is the concept of the d e t ermination of
attitudes b y the cognitive component, beliefs.
Triandis,
in a m o r e recent discussion
(1979), points out a basic
source of c o n t r o v e r s y in social psychology;
that o p e r a t i onalization and
m e a s u r e m e n t of a construct is dependent on h o w the construct
If a ttitude is linked to b e h a v i o r b y definition;
then,
is defined.
the behavioral
s c ientists conc e r n is to explore the conditions u n d e r w h i c h either a
s t r o n g or weak r e l a t i o n s h i p b e t w e e n verbal attitudes and b e h a v i o r are
likely to be observed.
C al d e r and Ross
tions of attitudes,
(1976),
conceive of attitudes as evaluative summaries of
u n d e r l y i n g beliefs.
attitudes,
in c o n s idering the psychol o g i c a l f o u n d a ­
A c c o r d i n g to this view,
in order to unde r s t a n d
it is first n e c e s s a r y to understand the information structures
or b e l i e f systems w h i c h u n d e r l i e attitudes.
Secondly,
it is n e c e s s a r y
to u n d e r s t a n d h o w these beliefs are p r o c e s s e d or integrated to prod u c e
the e v a l u a t i v e summary c a lled an attitude.
Some of the m a j o r contributions relevant to these issues
the c ognitive c o n s i s t e n c y a p p r o a c h of Heider's
include
(1946) b a l ance theory
w h i c h stresses the func t i o n of the p e r c e p t i o n of cons i s t e n c y in attitude
f o r m a t i o n and change.
A l o n g the same line of rea s o n i n g was Festinger's
3
(1957) t h eory of c o gnitive dissonance w h i c h examined perceptual
s istencies and modes of r e d u c i n g the r e s u l t i n g dissonance.
1972),
in c o n ­
Bern (1967,
in an a l t e r n a t e approach, p r o p o s e d a theory of self-perception
w h i c h suggested that b e h a v i o r might well be an antecedent rather than
a result of atti t u d e in that individuals
infer what their attitudes must
be from o b s e r v a t i o n o f what their b e h a v i o r is.
Thus,
there
is a vast
compl e x i t y of attitude constructs and a t t i t u d e - b e h a v i o r relationships
and a m u l t i p l i c i t y of a p p roaches taken in a t t e m p t i n g to gain an u n d e r ­
stan d i n g of attitudinal components and their processes.
The theoretical
a p p r o a ch focused u p o n in this study is that
d e v e l o p e d by F i s h b e i n
(1967)
and elaborated b y Fishbein and A j z e n
(1975).
This theoretical f r a m e w o r k has served to integrate m u c h of the currently
accep t e d a t t i t u d e - b e h a v i o r k n o w ledge into a theory that is explicit,
testable,
and w i d e l y g e n e r a l i z a b l e .
It is the application of this
theory in the area of attitudes toward a t t e n d a n c e / a b s e n t e e i s m that is
the subject o f this study.
The F i s h b e i n - A j z e n Model
R a t h e r t h a n p r o v i d i n g a simple d e f i nition of attitude,
and A j z e n
(1975) p r o p o s e d a conceptual
Fishbein
framework s y s tematically i n t e g r a t ­
ing t h e o retical attit u d e c o mponents and their und e r l y i n g processes.
T h e y n o t e that v a r i o u s
interpretations or definitions of "attitude"
h ave in the past implied d i fferent m e a s u r e m e n t procedures which c o n s e ­
q u e n t l y p r o d u c e d d i f f e r e n t results
in attitude studies.
The subsequent
co n f u s i o n as to e x a c t l y what "attitude" is has been the logical result.
F i s h b e i n and A j z e n
(1975) p r e f e r to define attitude b y its generally
a g reed u pon m o s t e s sential component w h i c h is the m a j o r characteristic
4
that d i s t i n g u i s h e s a t t i t u d e f r o m other constructs,
that is,
its evalu­
ative or affective nature.
A c c o r d i n g to this definition,
b e l i e f and attitude.
a disti n c t i o n mu s t be m a d e between
This d i s t inction implies the testable proposition
that beliefs and attitudes ha v e different d e t e rminants and that changes
in them can lead to d i f f e r e nt consequences.
Ajzen
(1975) use the term,
Accordingly,
" a t t i t u d e , ” to refer to affect,
toward or e v a l u a t i o n of an attitude object,
the same as R o k e a c h ’s
(1968)
i.e.,
feelings
and the t e r m "belief,” to
r epr e s e n t cogni t i o n or k n o w ledge about the object,
linkage of an object to some attribute.
Fishbein and
s p e cifically the
This definition,
d e f i n i t i o n of attitude,
while basically
emphasizes the
sepa r a t i o n of the concepts of attitude and belief.
The third g e n e r a l l y r e c ognized component of attitude,
al component,
behav i o r a l
the b e h a v i o r ­
is d i v ided in the F i s hbein-Ajzen model to refer to both
intentions and actions w i t h respect to or in the p r e sence of
the a ttitude object.
This disti n c t i o n between behavioral
intentions
and actual b e h a v i o r is m a d e since most theorists agree that attitudes
are c o ncerned w i t h p r e d i s p o s i t i o n s to behave r a t h e r than w i t h the
b e h a v i o r itself.
The F i s h b e i n - A j z e n
(1975) model,
then,
is a descr i p t i v e framework
of the r e l a t i o n s h i p b e t w e e n four b r o a d categories:
evaluation),
tions),
cognition
and b e h a v i o r
(opinions, beliefs),
(observed overt acts).
r e s e r v e d for only one of these categories,
conation
affect
(feelings,
(behavioral in t e n ­
The term "attitude" is
affect.
5
C o n ceptual F r a m e w o r k .of the Fish b e i n - A j z e n Model
The conceptual f ramework of the Fishbein - A j z e n model
for the r e l a ­
t i o n s h i p o f attitudes to b e h a v i o r can perhaps best be p r e s ented
schematically
(see
Figu r e 1}.
ance or n o n p e r f o r m a n c e of
A c c o r d i n g to this
framework, the p e r f o r m ­
a specific b e h a v i o r is d e t e rmined b y the
i n t e n t i o n to p e r f o r m that behavior.
Consequently,
the p r e d i c t i o n of
b e h a v i o r toward an
object f rom knowledge of attitude toward
object is a c c u r a t e
only insofar as that attitude influences the intention
to p e r f o r m the behavior.
beliefs,
This behavioral
that same
intention is a function of
not about the object of the behavior, but instead beliefs c o n ­
c e rned w i t h the b e h a v i o r itself.
A p e r s o n ’s atti t u d e toward p e r f o r m i n g a given b e h a v i o r is represented
as a f u n c t i o n of two types of beliefs.
the b e h a v i o r will
One of these is that p e r f o r m i n g
lead to c ertain consequences along w i t h h i s / h e r e v a l u ­
a t i o n of these consequences.
The other relevant beliefs,
j e c t i v e norms since they a r e n o r m a t i v e in nature,
labeled s u b ­
are beliefs that
c e r t a i n relevant others think that the p e r s o n should or should not
p e r f o r m the b e h a v i o r in question.
Subjective norms are combined multi-
p l i c a t i v e l y w i t h the s u b j e c t ’s m o t i v a t i o n to comply w i t h these norms.
A c c o r d i n g to this conceptual structure,
b u i l d i n g blocks,
o f attitudes,
then,
beliefs are the fundamental
an i n f ormational base that is the ultim a t e determinant
intentions,
and behavior.
The formation of attitudes,
is v i e w e d in terms of an information p r o c e s s i n g approach wherein
a p e r s o n ’s salient set of b e l iefs about the object determines h i s / h e r
a ttitude t o ward that object.
b e l i e f s as a whole,
A p p l i e d to behavior,
it is the set of
i n c l u d i ng b e h avioral intentions,
w h i c h are viewed
o
3
X*
33
3
P
O
pO
"d
■H
cd
cd
>
£
3 3
P X
O CD
33
P
3
£
O
P
cd
•rH
tin
PU
1
CQ
o
O
CD t
> 3
•H •H
P
O
•H
>
3
r3
CD
CQ
P 3
O P
CD CD
of
•
i
—>O
rC3 3
3 O
co U
/
Representation
CD
33
3
p
•H
P
P
<
2
to
<D
CD
3
Informational
I CD
P
<P
O
3
>
to w
CD
Xi O X
3
O
3
<p
•H
ip
3
o'
CD
to
3
O
CD
3
CD p <P
3 o O
o ' •H
CD > 3
to 3 O
3
CD O
03 U
P
3
bO
X •H
CD P
0Q 3
to
tp
CD
X
i t— \
t
o
O to
•H P •H P
p O CO CD O
CD -H
3S P
CQ > <P P
3 o O 3
O
CD r3
> CD to P H
•H CQ <p 3 P
CD 3 3
P
3 P ■H O >
S 3 iP •H H
P O CD <P p
O JQ 03 •H O
2 3 '— ' 3 S
J
X
p
pH
g
O
U
Schematic
tin
CD
CD
Fishbein-Ajzen
PS
0
pc
Model
of
Attitude-Behavior
Relationship
6
7
as a special case of beliefs,
that is the d e t e rminant of attitudes in
the a t t i t u d e - b e h a v i o r relationship.
The F i s h b e i n - A j z e n M odel of Attit u d e Formation
Fi s h b e i n and A j z e n
a ttitude formation.
(1975)
suggest an expectancy-value model of
The p r e d i c t i o n of behavioral
f u n c t i o n of the w e i g h t e d sum of two variables,
intentions is a
the attitude toward
p e r f o r m i n g the b e h a v i o r and the subjective n o r m as follows:
B 'v I =
(A )
+ (SN)
B''w1
v 'w2
w h e r e B is the behavior,
A
I is the intention to p e r f o r m the behavior,
is the a t t i t u d e towa r d p e r f o r m i n g the b e h a v i o r B, SN is the subjective
norm,
and w^ and w^ are e m p ir i c a l l y d e t e r m i n e d weights.
The attitude
toward p e r f o r m i n g a specific b e h a v i o r is pr o p o s e d to be a function of
the p e r c e i v e d c o n s equences of p e r f o r m i n g that be h a v i o r and of the
p e r s o n ' s e v a luations of those consequences:
A
B
=
n
L
b.e.
. . 1 1
i=l
w h e r e b is the b e l i e f that p e r f o r m i n g b e h a v i o r B leads to consequence
or outcome,
i_, e_ is the p e r son's e v a l u a t i o n of outcome i^, and n is the
n u m b e r of b e l iefs the p e r s o n s hold about p e r f o r m i n g be h a v i o r B.
The n o r m a t i v e component,
environment on behavior.
SN, deals w i t h the influence of the social
The s u b j ective n o r m is the person's p e r c e p t i o n
that p e o p l e who are important to h i m / h e r think he/she should or should
not p e r f o r m the b e h a v i o r in question.
A c c o r d i n g to Fishbein and Ajzen
8
(1975),
the general
s u b jective n o r m
(SN)
is dete r m i n e d b y the p e r ceived
e xpectations of s pecific r eferent individuals or groups,
and b y the
pers o n 's m o t i v a t i o n to c o m p l y w i t h those expectations:
n
I
b .m .
1-1
11
SN =
where b^ is the n o r m a t i v e belief, m^ is the m o t i v a t i o n to comply with
r eferent jl, and n. is the n umber of relevant referents.
T h ese two m a j o r d e t e r minants of behavioral
e mpirical weig h t s
intentions are given
in the p r e d i c t i o n e q u ation p r o p o r t i o n a l to their
r e l a t i v e importance.
Since adequate estimates of these weights for
e a c h individual are not generally available,
the accepted p r a c t i c e has
be e n to use m u l t i p l e r e g r e s s i o n techniques and standardized regression
c oefficients as estimates of the weights for the theory components.
The p r e s e n t v e r s i o n o f the theory,
in the form of a multiple r e g ression
equation i s :
i = (3b )ab ♦ ceSN)SN
where
I is the b e h a v i o r a l
intention.
The component of m o t i v a t i o n to comply can be expressed as the
person's
i n tention to c o m p l y w i t h the referent
d e t e r minants of
this inten tion
behavioral intention
m ^
—
I
and can
in question.
are the same as those of
any other
be expressed in equation form,
« (A„)
i (SN)
C v C J w^ v ' w 2
The
9
where m is the m o t i v a t i o n to comply w i t h the referent,
inte n t i o n to comp l y w i t h the refer e n t ,
A
1^, is the
is the attitude toward complyVj
ing w i t h the referent,
wi t h the referent,
SN is the subjective n o r m conc e r n i n g compliance
and w^ and
The F i s h b e i n - A j z e n
(1975)
are weights.
theory accounts for the influence on
intentions of additional v a r iables external to the model only through
their indirect influence on either of the two components
subjective norms)
Accordingly,
(attitude and
or on the relative weights of these components.
the attitude t oward the target object or p e r s o n will be
u n r e l a t e d to the b e h a v i o r a l
intention itself if it is not related to
e i ther the attitudinal or n o r m a t i v e component of behavioral
intention.
O n l y if the component v a r i a b l e in qu e s t i o n carries a significant weight
in the r e g r e s s i o n e q u a t i o n p r e d i c t i n g behavioral
intention will attitude
t o ward the object be r e l a t e d to or p r e d i c t i v e of intentions.
b e havioral
i ntention w h i c h is considered to be the determinant of overt
volit i o n a l behavior.
(1977)
It is the
A n u m b e r of studies are cited b y Ajzen and Fishbein
and F i s h b e i n and A j z e n
(1975)
in support of this theoretical
formulation.
F i s h b e i n and A j z e n ’s S p e c i f ic a t i o n s for M e a s u r e m e n t
A m a j o r fact o r in the p r e d i c t i o n of overt b e h a v i o r from behavioral
i ntentions is the n e c e s s i t y for correspondence with respect to the
level of s p e c i f i c i t y b e t w e e n intentions and be h a v i o r and also between
intentions and the compo n e n ts of intentions.
F i s h b e i n and Ajzen
(1975),
in t h e i r devel o p m e n t o f a f r a mework for linking attitudes and behaviors,
are h i g h l y s pecific as to the m e a s u r e m e n t methods that they consider
appropriate.
Consequently,
any study d e s i g n e d to investigate this model
10
risks the p o s s i b i l i t y o f t e s t i n g constructs other than those designated
if other m e t h o d s of m e a s u r e m e n t are employed.
The p r o c e d u r e recommended
b y F i s h b e i n and A j z e n is to m e a s u r e attitude "by a p r o c e d u r e which
locates the subject on a b i p o l a r affective or evaluative dimension
v i s - a - v i s a given object"
(Fishbein § Ajzen,
1975, p.
11).
is an exact d e f i n i t i o n of the semantic-differential scale
§ Tannenbaum,
This p r o c edure
(Osgood,
Suci,
1957).
In contrast to the e v a l u a t i v e natu r e of attitudes, beliefs represent
information.
Diff e r e n c e s a m o n g individuals in this respect are defined
in terms of b e l i e f s t r e n g t h or the p e r c eived likelihood that an object
has or is a s s o c i a t e d w i t h a p a r t i c u l a r attribute.
p r o c e d u r e for the m e a s u r e m e n t of b e l i e f strength,
The r e c o m m e n d e d
then, p l a c e s the s u b ­
ject a l o n g a d i m e n s i o n of s u b j ective p r o b a b i l i t y i n v olving an object and
some related attribute.
ceiv e d to possess,
For example,
the mo r e mon e y a p e r s o n is p e r ­
the h i g h e r should be the subjective p r o b a b i l i t y that
the p e r s o n is wealthy.
Since b e h a v i o r a l
beliefs,
i n t e n tion is conceptualized as a special case of
the s t r e n g t h of a b e h avioral
intention is app r o p r i a t e l y
m e a s u r e d b y a p r o c e d u r e w h i c h places the subject along a subjective
p r o b a b i l i t y d i m e n s i o n i n v o l ving a r e l a t i o n b e t w e e n h i m s e l f / h e r s e l f and
some action.
For example,
the s t r ength of an intention to attend church
on S u n d a y w ould be m e a s u r e d b y the s u b j e c t ’s p r o b a b i l i t y r a t i n g of the
concept,
"I will attend c h u rch S u n d a y , ” on scales anchored b y "probable-
i m p r o b a b l e ” or " a g r e e - d i s a g r e e ."
A c c o r d i n g to F i s h b e i n and A j z e n
(1975)
only a limited n u m b e r of
salient beli e f s can serve as de t e r m i n a n t s of attitude at any given time.
11
Therefore, m e a s u r e s of b e l iefs should also include assessment of their
s a l i e n c y in the subject's b e l i e f hierarchy.
This can be a c c o m p l i s h e d
by c o n s i d e r i n g as salient only the first few responses elici t e d in a
f r e e - r e s p o n s e format w h e n subjects are asked for a listing of c h a r a c ­
teristics,
qualities,
and attributes p o s s essed by an object or for the
co n s e quences of p e r f o r m i n g a behavior.
This p r o c e d u r e is s i m ilar to content analysis.
o f b e l iefs for a p o p u l a t i o n is to be identified,
When a mod a l set
the responses obtai n e d
f r o m a r e p r e s e n t a t i v e sample are first grouped according to similarity
and the f requency of each similar b e l i e f is counted.
It is n e c e s s a r i l y
a m a t t e r of judgment as to w h e t h e r or not similar - a p p e a r i n g beliefs are
to be consi d e r e d identical.
The set of beliefs to be included in the
modal set is then a r b i t r a r i ly set as the n u m b e r of beliefs
to account for a stated p e r centage,
b y the sample.
such as 75%, of all beliefs emi tted
This p r o c e d u r e is d e s c ribed by Aj z e n and Fishb e i n
By definition,
Ajzen
overt behaviors are observable acts.
(1975) v i e w b e h a v i o r as c o n s i s t i n g of four elements:
target,
sufficient
situation,
types of b e h a v i o r a l
and time.
(1980).
F i s hbein and
behavior,
T h e y make a distinction b e t w e e n different
crite r i a in terms of the variance of the criteria
w i t h r es p e c t to one or m o r e of these elements.
A single-act c r i t e r i o n , i.e.,
act,
the single observation of a single
is always s pecific w i t h respect to the four elements of behav i o r
as it involves a d i r e c t l y o bservable r e s p o n s e to a specific target,
a given situation,
at a given point in time.
in
A repeated-observation
c r i t e r i o n is an index of b e h a v i o r d e r i v e d from repeated observations of
the same behavior,
such as observations across several trials in an
12
experiment.
Such c riteria can represent generalizations across targets,
across situations,
a behavioral
or across time.
A m u l t i p l e - a c t criterion represents
index compu t e d from observations of different behaviors
w i t h respect to a given target,
the same p o i n t
in time.
in a given situation,
For example,
at a p p r oximately
withdrawal behav i o r in a social
s i t u a t i o n can be m e a s u r e d b y the degree of conversational participation,
eye contact, phys i c a l distance,
group.
and the amount of time spent w i t h a
A c o m b i n a t i o n of r e p e a t e d observations of m o r e than one b e h a v i o r
w o u l d be c o n s i d e r e d a m u l t i p l e - a c t , rep e a t e d - o b s e r v a t i o n c r i t e r i o n .
Since b e h a v i o r a l obser vations are data,
such observations can be
subject to the same p r o b l e m s of u n r e l i a b i l i t y and invalidity as any other
form of data.
For this reason,
Fishbein and Aj z e n
(1975) note that
r igor o u s analyses of b e h a v i o ral data are essential for an u n d e r s t a n d i n g
o f the r e l a t i o n b e t w e e n attitude and b e h a v i o r and that inconsistent
r e s e a r c h findings f rom a t t i t u d e - b e h a v i o r studies m a y be due to the use
of inapp r o p r i a t e b e h a v i o r a l measures.
They conclude that,
in reg a r d to
s i n gle-act criteria:
Not e v e r y b e h a v i o r w i t h respect to some object
the a t t i t u d e toward that object
r e p e a t e d - o b s e r v a t i o n c r i teria
is related to
[and that] multi p l e - a c t and
. . . when p r o p e r l y constructed
on the b a s i s of stan d a rd scaling p r o c e d u r e
. . . can serve as
indicants o f attitude.
1975, pp.
(Fishbein § Ajzen,
356-357)
The B e n tler and Speckart Model
B cntlcr and Speckart
(1979) p r o p o s e d and tested a m o d i f i c a t i o n and
e x t e n s i o n o f the F i s h b e i n and Ajzen model.
model,
In the Bentler-Speckart
w h i c h they cons i d e r to be "the most t h e o r e t i c a l l y adequate
13
c a u s a l - p r e d i c t i v e s y s t e m r elevant to a v a r i e t y o f behavioral domains"
(1979, p. 455),
affect
(attitude)
has a direct effect on behavior in
addition to its indirect influence on
ence on intention.
A
second
b e h a v i o r b y means of its inf l u ­
m o d i f i c a t i o n is the addition of previous
b e h a v i o r to the m o d e l .
This previous b e h a v i o r is p o s t u l a t e d to have an
effect on b o t h current
intentions and on future be h a v i o r that cannot be
acco u n t e d for b y the original Fishbein and Ajz e n model.
of these two approaches
A schematic
is p r e s e n t e d in Figure 2.
Bentler and Speckart
(1979) r e a s o n that,
is conscious and t h e r e b y cognitive in nature,
since behavioral
intention
the Fish b e i n - A j z e n theory
wh i c h p r o p o s e s that affect impacts be h a v i o r only b y means of the r e g u ­
lation o f intention or p r e m e d i t a t i o n
(conation-cognition)
intuitive in m o s t domains of behavior,
and has
is c o u n t e r ­
less accurate and generalizable,
less p r e d i c t i v e p o w e r than their conceptualization.
Th e y also p r o p o s e that the role of previous b e h a v i o r in a c c o unting
for future b e h a v i o r is i n a d eq u a t e l y m o d e l e d in the Fishbein - A j z e n
ap p r o a c h of indirect influence through attitudes and s u b j ective norms.
Bent l e r and S peckart state that previ o u s b e h a v i o r m a y "circumnavigate
these factors
in its c a u s a t ion of subsequent b e h a v i o r in the same w a y
that attitudes c i r c u m n a v i g a t e intentions"
oretical formulation,
(1979, p. 454).
This t h e ­
w h e r e b y b e h a v i o r has an independent role in the
p r e d i c t i o n o f future behavior,
f o r m u lations such as Bern's
is consistent w i t h other theoretical
(1967,
1972)
s e l f - p e rception theory in w h i c h
a t titudes m a y be g e n e r a t e d from self-perceptions of behavior.
r e s e a r c h findi n g s
Other
i n d i c a t i n g a relationship b e t w e e n past and subsequent
b e h a v i o r w o u l d a p pear to be consistent wi t h b o t h the direct and indirect
14
Schem a t i c R e p r e s e n t a t i o n of F i s h b e i n - A j z e n Model
Attitude
toward
Behavior
Behavioral
Intentions
Target
Behavior
Subjective
Norms
Schematic R e p r e s e n t a t i o n of B e n tler-Speckart M o d i f i c a t i o n
Previous
Behavior
Behavior
Intentions
Attitude
toward
Behavior
Subjective
Norms
Figure 2
Target
Behavior
15
p r e v i o u s - b e h a v i o r influ e n c e models.
In the B e n t l e r - S p e c k a r t
the same data:
(1979)
study,
three models were tested with
the two schematically re p r e s e n t e d in Figure 2 and an
i n t e r mediate m o d i f i c a t i o n of the Fishbein-Ajzen model.
S peckart
(1979)
Bentler and
c o l l e c t e d their data fol l o w i n g Fishb e i n and A j z e n 1s
r e c o m m e n d e d a p p r o a c h of u s i n g s e m antic-differential
scales.
A sample of
228 college students were a sked three questions on each of the constructs
o f attitudes,
subje c t i v e norms,
and intentions at one point in time.
B ehavior was m e a s u r e d twice with i n a t w o -week period.
tions of the tests for each model,
these five variables were assessed
f o r each of t h r e e atti t u d i n al domains:
d r u g use.
To obtain r e p l i c a ­
alcohol,
marijuana,
and hard
The b e h a v i o r m e a sures were not observations of b e h a v i o r but
r a t h e r q u e s t i o n n a i r e r e s p o n ses of self-report of b e h a v i o r for the twow e e k p e r i o d p r i o r to the time of d a t a collection.
The first b e h avior
m e a s u r e was taken at the time of the complete que s t i o n n a i r e a d m i n i s t r a ­
tion;
the second two weeks
later.
The first analysis was a c o m p arison of the Fish b e i n - A j z e n model as
shown in the u p p e r h a l f of Figure 2 w i t h a m o d i f i c a t i o n w h i c h added
o n l y a direct p a t h f r o m attitude to b e h a v i o r to the Fishbein - A j z e n
model.
Thus,
this m o d i f i e d model did not include p r i o r behavior.
A t t i t u d e s and s u b j e c t i v e norms were the exogenous or independent v a r i ­
ables;
intentions
variables.
and s u b s e quent b e h a v i o r were dependent or endogenous
E n d o g e n o u s v a r i ables are d e f i n e d as v a r i ables whose causes
are c o m p l e t e l y d e t e r m i n e d b y v a r i ables included in the causal m o d e l ;
exogenous v a r i a b l e s are d e t e r m i n e d b y causes lying outside the model.
F or the additional
c o m p a r i s o n of models with the f u l ly-expanded
16
B e n t l e r - S p e c k a r t model,
as p r e s e n t e d in the lower h a l f of Figure 2; the
m e a s u r e s o f p r e v i o u s b e h a v i o r were included.
Thus, pr i o r behav i o r is
c o n c e i v e d as a latent v a r i a b l e wh i c h was included in the causal model.
A fact o r analysis of the data supported the c o n c l u s i o n that the
th e o retical constructs h y p o t h e s i z e d as latent factors were adequately
a s s e s s e d and were r e a s o n a b l y indicated by the ob s e r v e d variables.
Para m e t e r s
for each causal model
first m o d i f i e d model,
(Fishbein-Ajzen model,
Bentler-Spe ckart
and B e n tler-Speckart full y - e x p a n d e d model)
e s t i m a t e d b y a c o m p u t e r program,
LISREL IV (Joreskog § Sorbom,
H i e r a r c h i c a l m o d e l s w e r e generated by adding p a r a m e t e r s
to or r e m o v i n g t h e m from the mode l s b e i n g tested.
were
1978).
(causal paths)
Chi-square d i f f erence
tests of the null h y p o t h e s i s that each par a m e t e r in q u e s t i o n is not
p r e s e n t in the p r o p o s e d causal structure in the p o p u l a t i o n were u s e d to
compare c o m p e t i n g models.
Results of the statistical analysis supported
B e n t l e r and Speckart's
(1979) h y p othesis that the a d d ition of three
structural parameters,
i.e.,
direct paths from a t t itude to subsequent
b e h a v i o r and from p r e v i o u s b e h a v i o r to b o t h intention and subsequent
behavior,
is n e c e s s a r y for the causal model
to su c c e s s f u l l y reproduce
the data.
The Bentler-Speckart
executed;
however,
(1979)
study was c a r efully conceived and
some comments on it appear to be in order.
latent v a r i a b l e s a s s e s s e d in their study
p r i o r behavior,
intention,
(attitude,
and target behavior)
The five
subjective norms,
were each assesse d by
an overall m e a s u r e r a t h e r than b y separate m e a s u r e s of their components.
A t t i t u d e w a s m e a s u r e d as the evaluative component only and did not
include b e l i e f s and b e l i e f strength.
Subjective norms were m e a s u r e d by
17
b e l i e f s t r e n g t h but did not encompass the m o t i v a t i o n to comply.
Be h a v i o r was m e a s u r e d b y s elf-report rather than b y observations of
o vert behavior.
The element of response bias inherent in self-reports
o f s o c i a l l y - c e n s u r e d or p o t e n t i a l l y illegal acts should be considered
since the target beha v i o r s of the study were the use of alcohol,
marijuana,
and hard drugs.
Thus,
the measu r e m e n t p r o c e d u r e did not
c o r r e s p o n d to F i s h b e i n and A j z en's recommendations and m a y have undu l y
b i a s e d the r e s ults a g a i n s t the Fishbein and Ajzen model.
The s t u d y r eported here is a comp a r i s o n b e t ween the Fishbein-Ajzen
model and the B e n t l e r - S p e c k a r t f u l ly-expanded model,
ferences in p r e d i c t i v e p o w e r b e t w e e n them.
to s h o w no significant
testing for d i f ­
If the expanded model were
increase in p r e d i c t i o n over the basic Fishbein-
Aj zen m odel and if the additional
causal paths h y p o t h e s i z e d in the
B e n t l e r - S p e c k a r t model do not d e m o n s t r a t e significant structural p a t h
coefficients,
then support for the Fishb e i n - A j z e n model w o u l d seem w a r ­
r a n t e d on the basis of the m o s t p a r s i m o n i o u s e x p l a n a t i o n of the data
and the r e l a t i o n s h i p s represented.
Method
Subjects
Subjects were 259 college students,
of b o t h sexes, who were
e n r o l l e d in the 1980 summer session at the U n i v e r s i t y of N e b r a s k a at
O m a h a in p s y c h o l o g y classes w h i c h met five days p e r week.
E x t r a credit
t o w a r d the student's course grade was given for comp l e t i o n of the
questionnaire.
one class,
data.
For the final da t a analysis,
23 subjects,
comp r i s i n g
were e l i m i n a t e d due to insufficient behavioral atte n d a n c e
This left a final
sample of 236 for the p r i m a r y study sample.
18
P r ocedure
D e m o g r a p h i c d a t a were collected on each subject in a pilot sample
and in the p r i m a r y study sample.
sex,
age, m a r i t a l status,
These data included information on
student status
(number of hours
in which
c u r r e n t l y enro l l e d and expected to be enrolled in the coming fall
semester),
expe c t e d date of graduation,
employment status
(number of
hours w o r k e d p e r w e e k c u r r e ntly and expec t e d in the fall),
income,
U.S.
citizenship,
and race.
family
The purp o s e of this data collection
was to assess the r e p r e s e n t ativeness of the pilot sample relative to
the p o p u l a t i o n f r o m w h i c h the p r i m a r y study sample was also drawn.
Data on e x p e c t e d fall semes ter s c h ool/work status and on the expected
d ate of g r a d u a t i o n were collected b e c a u s e of the p o s s i b i l i t y that these
v a r i a b l e s m i g h t serve a m o d e r a t i n g function.
A pilot sample of 123 summer school students in p s y c h o l o g y classes
was u s e d to i dentify m o d a l beliefs of the consequences of the target
behav i o r s
(class atten d a n c e /absenteeism)
r e l a t i o n to these behaviors.
Fishbein
(1980)
and the significant others
in
The p r o c e d u r e outlined b y A j z e n and
and p r e v i o u s l y d e s c r i b e d was followed.
Subjects were
asked in a f r e e - r e s p o n s e format to list the beliefs that came to m i n d
as p o s s i b l e c o n s e quences of the b e h a v i o r of a t t ending or of b e i n g absent
from class.
T h e y were also asked to identify individuals or groups
w h o s e opinions w i t h respect to these behaviors are important to them.
See A p p e n d i x A for the p i l o t sample questions.
T a b u l a t i o n of r e s ults from the 123 returned question n a i r e s yielded
1,093 r e s p o n s e s to the b e l i efs
O f these,
(behavioral consequences)
questions.
90 resp o n s e s w e r e d i r e c t e d at the consequences of attending
19
summer school p e r se and an additional 97 r e s p onses were "none" or
blanks.
E l i m i n a t i o n of these null or nonre l e v a n t responses
left 906
a t t e n d a n c e / a b s e n c e b e h a v i o r al consequences w h i c h were then grouped and
t a b u l a t e d as categories.
E leven response categories represented 698
r e s p o n s e s or 77% of total relevant responses elicited.
A n arbitrary
d e s i g n a t i o n of 75% of total responses had b e e n es t a b l i s h e d as the
c r i t e r i a for salient m o d a l responses to be r e t a i n e d for the p r i m a r y
study q u e stionnaire.
T hese ele v e n response categories were then used
in formu l a t i n g c o r r e s p o n d i n g questions to be r a t e d by subjects
p r i m a r y study
in the
(see A p p e n d i x B for the p r i m a r y study q u e s t i o n n a i r e ) .
Q u e s t i o n 6 of the p i l o t study q u e s t i o n n a i r e was designed to elicit
modal refer e n t s
for the population.
this q u e s t i o n was 260.
alone" and 4 were
categories
Of these,
"no one."
included 157
(83%)
67
The total n u m b e r of responses to
(25.77%)
were
"myself" or "myself
O f the r e m a i n i n g 189 responses,
of the total.
five
These were the response
categories u s e d in the p r i m a r y study as mo d a l referents.
The p r i m a r y study u s e d students in nine,
classes.
five-week summer school
C o m p l e t e d a t a were obtained from 236 subjects out of 295
s tudents o r i g i n a l l y r e g i s t e r e d for these classes.
Of these 295 students,
25 e i t h e r p r o v i d e d n o a t t e n d a n c e d a t a or p r o v i d e d n o n e after the first
w e e k of a t t endance d a t a collection.
If the assu m p t i o n is made that
these students d r o p p e d out of the classes,
the v o l u n t e e r rate was 87.4
for p a r t i c i p a t i o n in the study.
The target b e h a v i o r or b e h a v i o r of interest was class attendance/
absence w h i c h was a s s e s s e d b y p a s s i n g attendance sheets
to sign.
for students
Students were u n i f o r m l y told at the b e g i n n i n g of the session
20
that class a t t e n d a n c e was not a factor in grading,
but that attendance
sheets w o u l d be p a s s e d for "administrative purp o s e s . "
In order to avoid
c o n f o u n d i n g r e g u l a r class attendance with the special case of attending
class on a test day,
no a t tendance was collected on test days.
class there were some days, u n s y s t e m a t i c a l l y distributed,
atten d a n c e was not collected.
In each
on which
Because of the unequal n u m b e r of data
c o l l e c t i o n days b e t w e e n classes,
the m e a sures of b e h a v i o r were the ratios
of days a t t e n d i n g to the n u m b e r of days in that time period.
Prior b e h a v i o r was the a t t e n d a nce/absence data collected for two
weeks b e f o r e the c o l l e c t i o n of attitude,
data.
subjective norms,
and intention
The o nly e x p l a n a t i o n given for collection of these attendance
d a t a was that t hey w e r e "for a d m i nistrative p u r p o s e s . "
of the s u mmer school session,
At the midpoint
(i.e., b e t w e e n weeks 3 and 4),
a semantic-
d i f f e r e n t i a l q u e s t i o n n a i r e was administered to all subjects to assess
a ttitudes t o ward the targ e t behavior,
intentions.
subjective norms,
and behavioral
T a rget b e h a v i o r was the atten d a n c e / a b s e n c e d a t a obtained
after c o l l e c t i o n of the q u e s t i o n n a i r e data and until the end of the
summer school class session,
the final two weeks of the session.
The
first w e e k of the f i v e - w e e k session was omitted from the study as that
is t y p i c a l l y a w e e k of instability in wh i c h m a n y students disenroll
from t h e i r courses.
Measures
M u l t i p l e i n d icators
(measures) were used to assess the r e l i a b i l i t y
of m e a s u r e m e n t for each latent va r i a b l e
(construct)
and to remove
m e a s u r e m e n t e r r o r from the relatio n s h i p s among the latent variables.
Thus,
beha v i o r a l m e a s u r e s were combined into several pre- and
21
p o s t - i n t e n t i o n periods.
The ele v e n days of class p r i o r to and including
the day of q u e s t i o n n a i r e a d m i n i s t r a t i o n were d e s i g n a t e d as p r i o r behavior
m e a s u r e s and d i v i d e d into three periods.
Likewise,
the post-que s t i o n n a i r e
p e r i o d of eight days was d e s i g n a t e d as target b e h a v i o r and divided into
two periods of f o u r days each.
As stated earlier,
these measures were
the ratio of days a t t e n d e d to days attendance was taken.
I ntention was a s s e s s e d b y two questions,
class every s e s s i o n
some times
. . ."
"I intend to attend this
. . .," and "I intend to be absent from this class
Scales w e r e anchored w i t h "likely" and "unlikely" at
their respe c t i v e endpoints with rever s e d s c o ring for the intention to
be absent question.
Two m e a s u r e s
of the a t titude toward the b e h a v i o r were used.
was a scale score d e r i v e d f rom the s u m mation of seven items
in the q u e s t i o n n a i r e
One
(item 3
in A p p e n d i x B) e v a luating the b e h a v i o r of class
a t t e n d a n c e on seven - p o i n t scales w i t h endpoints a n c hored b y "importantu n i m p o r t a n t , " " w o r t h l e s s - v a l u a b l e , " "good-bad," "rewarding-p u n i s h i n g , "
and so forth.
The o t h e r m e a s u r e of attitude was deri v e d from the rat i n g
of each of e l e v e n c o n s e q u e n ces of the be h a v i o r
the q u e s t i o n n a i r e in A p p e n d i x B ) .
the p i l o t sample;
(items 5 t h r o u g h 15 of
These consequences were obtained from
the r a t i n g given to each c o n s equence on the seven-;
point scale was m u l t i p l i e d b y a specific eva l u a t i o n of it
(items
t h r o u g h 11 in the e v a l u a t i o n s e c tion of the q u e s t i o n n a i r e ) .
1
These eleven
p r o d u c t s of b e l i e f s t r e n g t h and e v a l uation w e r e summed to form a score.
S u b jective norms were also assessed by two measures.
was a general s u b j e c t i v e n o r m measure,
The first
"Most p e o p l e w h o are important
to m e think I should a t t e n d this class every d a y d u r i n g the summer
22
session.”
The r a t i n g given to this question was m u l t i p l i e d b y a rat i n g
of m o t i v a t i o n to c o mply w i t h this per c e i v e d b e l i e f whi c h was obtained
b y asking,
" G enerally speaking,
h o w m u c h do y o u want to do what others
w h o are important to y o u think y o u should d o ? ”
The other subjective
n o r m m e a s u r e was obtai n e d from ratings given to p e r c e i v e d beliefs of
s p e c i f i c signi f i c a n t others,
such as parents,
so f orth o b t a i n e d f r o m the pilot sample
friends,
and
(items 3 through 7 in the q u e s ­
t ionn a i r e s e c t i o n on '-how y o u think other p e o p l e wou l d
b e h a v e ” ).
instructors,
like y o u to
These r a t i n g s w e re then m u l t i p l i e d by the subject's m o t i v a t i o n
to c o mply w i t h these p e r c e i v e d b e l iefs
p a g e of the questionnaire)
(items 16 through 2 0 on the last
and the products summed.
Analysis
The m e t h o d o f data analysis us e d in this study was an e x t ension of
p a t h a nalysis d e v e l o p e d b y J o r e s k o g and Sorbom
(1978)
in an effort to
combine the e f f i c a c y of p a t h analysis in e x p l i c a t i n g u n d e r l y i n g r e l a ­
t i o n s h i p s w i t h a c o n f i r m a t o r y factor analytic ap p r o a c h used to identify
the factors
(latent variables)
in such relationships.
Whi l e a pa t h
a n a l y t i c a p p r o a c h is a p p r o p r i a t e for theory t e s t i n g and cla r i f i c a t i o n
(Billings § Wroten,
1973;
Li,
1978;
C o o k $ Campbell,
1975; N amboodiri,
Carter,
c e r t a i n limitations and s h o rtcomings
c u s s i o n of p a t h a n a l y s i s ) .
Sorbom
(1978)
1979;
§ Blalock,
1975), p a t h analysis has
(see A p p e n d i x D for a general d i s ­
The technique dev e l o p e d b y J o r e s k o g and
e l i minates or avoids m a n y of these p r o b l e m s and has been
u s e d in r e c e n t studies e x p l o r i n g causal models
M c Garvey,
K e r linger § Pedhazur,
1980;
Pedhazur, N o t e 1).
Also,
(e.g., M a r u y a m a §
this m e t h o d was deemed
e s p e c i a l l y a p p r o p r i a t e since it was used b y Bent l e r and Speckart
(1979,
23
1981)
in t h e i r tests o f m o d i f i c a t i o n s of the F i s h b e i n - A j z e n model.
This m e t h o d is an a nalysis of linear structural relationships by the
m e t h o d of m a x i m u m like l i h o o d o p e r a t i o n a l i z e d in the computer p r o g r a m
L I S R E L IV.
M u l t i p l e o b s e r v ed indicators of u n o b served latent constructs
are u s e d to infer r e l a t i o n s hips b e t w e e n the latent, u n m e a s u r e d variables.
This a nalysis provi d e s a m e a s u r e m e n t model and a causal model.
The
i n clusion o f m e a s u r e m e n t error or u n i q u e varia n c e as explicit parameters
in the m odel p e r mits causal regr e s s i o n parameters to be e s t i m a t e d w i t h ­
out the influence of m e a s u r e m e n t error.
The F i s h b e i n - A j z e n and B e n t l e r - S p e c k a r t models were compared by
e s t i m a t i n g the v a r i o u s structural par a m e t e r s of a saturated model,
a m o del
in w h i c h all p aths p o s s i b l e were estimated,
estim a t e s
i.e.,
and then c o m p uting
for p a r a m e t e r s of nes t e d m o d i f i c a t i o n s of this sat u r a t e d model.
This w a s a c c o m p l i s h e d b y s e t t i n g v a r i o u s p a r ameters equal to zero, that
is, r e m o v i n g p a t h s f rom the model.
Two c o m p a r a t i v e indices w e r e used.
First,
the critical ratio of a
speci f i c p a r a m e t e r gives a signif i c a n c e test for that parameter.
The
critical r a t i o of each causal p a r a m e t e r r e p orted as a t-value is the
r atio o f the u n s t a n d a r d i z e d LISREL p a r a m e t e r estimate d i v ided b y its
st a n d a r d error.
These ratios,
due to the large sample size,
p r e t e d as s t a n d a r d normal d evia t e s and represent
for the parameter.
At an a lpha level of
are inter­
levels of significance
.05, a critical ratio of less
than 1.96 w o u l d indicate the n o n - s i g n i f i c a n c e or e x p e n d a b i l i t y of the
speci f i c parameter.
In addition,
a c h i - s q u are d i f f e r e n c e test b e t ween the hierarchical
m o d e l s t e sted the null h y p o t h e s i s that the specific p a r a m e t e r omitted
24
in the n e s t e d m odel
tion.
1980;
is not p res e n t
in tbe causal structure in the p o p u l a ­
The r a t i o n a l e for the chi-square d i f ference test
Bentler § Speckart,
1979,
1981)
(Bentler § Bonett,
is b a s e d on goodness-of-fit chi-
square tests and the a s s o c i a t e d degrees of f r e e d o m for each model compared.
In the case of p a r a m e t e r nesting,
the m o d e l s to be compared differ only in
that the p a r a m e t e r v e c t o r of the m o r e r e s t r i c t e d model
is a special case
of the p a r a m e t e r v e c t o r of the less r e s t r i c t e d model, w i t h certain p a r a ­
m e t e r s c o n s t r a i n e d to e q u a l i t y or known constants.
of model
The null hypo t h e s i s
is
e q u i v a l e n c e and the difference b e t w e e n chi-squares w i t h degrees
of f r e e d o m equal to the d i f f e r e n c e in para m e t e r s estimated provides a
stat i s t i c a l test of that null h y p o thesis and of the statistical n e c e s s i t y
of the p a r a m e t e r s that d i f f e re n t i a t e the models.
Results
Pilot and P r i m a r y Study Sample C o m p arisons
Analyses
of d e m o g r a p h i c data ob t a i n e d from subjects in the pilot
and p r i m a r y study samples s u pported the h y p othesis that the two samples
were d r a w n f rom the same population.
m e a s u r e d d i f f e r e d at the
O n l y two of the fifteen v a r iables
.05 level of significance;
income and e x p e c t e d y e a r o f graduation.
report of family
The p r i m a r y study sample
r e p o r t e d an e a r l i e r date o f graduation and a greater p r o p o r t i o n reported
lower family income.
All of the comparisons are re p o r t e d in Table 1 in
A p p e n d i x C.
M e a s u r e m e n t Model
The m e a s u r e m e n t model p rese n t s the standa r d i z e d factor loadings of
the m e a s u r e d v a r i a b l e s on the latent factors.
lambdas
These
loadings,
or
(A), m a y be i n t e r p r e ted as v a l i d i t y coefficients r e f l e c t i n g the
25
degree to w h i c h the o b s e r v e d v a r i a b l e s adeq u a t e l y m e a s u r e the specified
un d e r l y i n g construct.
These para m e t e r s range from
r e p o r t e d b y B e n t l e r and S peckart
(1979)
.609 to 1.0, a range
as adequate.
r e presents the p r o p o r t i o n of the variables'
Unique variance
va r i a n c e that is not
ac c o u n t e d for b y the fact o r s and includes m e a s u r e m e n t error.
given b y e psilons
(e) and deltas
(Figures 3 and 4).
This is
(6 ) in the schematic representations
The top h a l f of Tables C2 and Cll in A p p e n d i x C
p r e s e n t s the m e a s u r e m e n t m odel for each causal
structure tested.
Structural Model
The s t r uctural or causal model estimates para m e t e r s of the r e l a ­
tions h i p s b e t w e e n latent,
removed.
to
Gammas
u n m e a s u r e d var i a b l e s w i t h error of meas u r e m e n t
(y) are i n t erpreted as path coefficients from exogenous
e n d o g e n o u s variables;
b etas
b e t w e e n e n d o g e n o u s variables.
(3 ) are interpreted as p a t h coefficients
The r e l ationships b e t w e e n latent exogenous
varia b l e s are given b y the phis
v a r i a b l e s are r e p r e s e n t e d b y zis
($).
(£).
Residuals of latent endogenous
The b o t t o m h a l f of Tables C2
t h r o u g h Cll p r e s e n t s the s t ructural model for each causal structure
tested.
The f o l l o w i n g s u m m a r y of results is ba s e d on the data p r e s e n t e d
in these tables
in A p p e n d i x C.
Mo del C o m p a r i s o n s
Figure 3 p r e s e n t s the structural models of relationships b e t w e e n
attitudes,
s u b j e c t i v e norms,
b e h a v i o r is not included.
intentions,
and target behavior.
Circles r e p r esent latent,
Prior
unmeasured v a r i ­
ables and r e c t a n g l e s r e p r e s ent the observed, m e a s u r e d variables.
D o u b l e - h e a d e d arrows repre s ent c o v ariance and sin g l e - h e a d e d arrows
r e p r e s e n t h y p o t h e s i z e d causal paths.
Figure 4 pr e s e n t s the structural
26
ta
•H
(A
X
r—I
P
5
-4
W
c£
co
p
o
4h
Pd
0
T-(
0
rQ
P
-4
,—
Pd
0
Pd
P
r-1
o
X
w
P
O
•H
>
P
,P
0
04
P
O
•H
P
Oh
'—'
r-H
0
DO
P
•H
Pd
bfl P
p •H
•H m
Pd
p rP
•H -P
P
PH
-P
■P P
cd P
Ph P
o
•P •rH
CH
P
cd •H
o P
•H bO
cp •H
•H (A
P P
fcuO o
•H p
tA
p
cd
0
0 p
P p
cd o
o •H
•H pd
Pd P
P •rH
•H
pd
0
s
p
0
tsi
•r->
<1
p
•H
0
rO
X
tA
•H
H,
to
CD
u
P
bfl
•H
P-,
<A
U) 0
<D £
P -H
•H i—I
r“H
P
Pd <D
•H ^
i—I o
O U
CO CQ
27
CO
X
ctS
i—t
G
<
04
OiS
co
00
G
•H
h3
0
i-H
0
j=>
cj
,-q
/—\
0
r— 1
o
g
i—i
p
o
•H
>
Pj
A
0
CQ
00
c
•H
rt3.
G
•H
4-1
G
•H
4-1
JG
P
pi
04
03
+J P
p3 G
04 G
o
+J •H
G 4-1
pj •H
o G
•H bO
4H •H
•rH to
c G
bO O
•H c
CO
0
O
•H
G
Cu
'—
1
i—t
0
h3
O
S
0
p
G
O
o •H
•H T i
T3 c
G •H
•H
+->
p3
-
CO
CD
C
•H
t
to
CD
C
-H
i—i
i—I
g
h3 0
•H ^
i-i O
O f-t
CO CQ
>- o
28
m o d e l s test e d w i t h p r i o r b e h a v i o r included.
E x o g e n o u s variables,
defined as v a r i a b l e s whose causes
of the h y p o t h e s i z e d model,
prior behavior
are attitude
lie outside
(A), s u b j ective norms
(PB) w h i c h is added in Figure 4.
(SN), and
E n d o genous variables,
d e f i n e d as those w hose causes are c o m p letely d e t e rmined w i t h i n the
h y p o t h e s i z e d model,
are intention
(INT)
and target b e h a v i o r
T h e m o d e l s c o r r e s p o n d i n g to the F i s h b e i n - A j z e n
(1975)
(TB) .
formulation,
i.e.,
those w h i c h did n o t include p r i o r b e h a v i o r data are labeled FA-1,
FA-2,
and FA-3
(see Table 1).
p r e s e n t e d in Figure 3.
The p a r a m e t e r s are sch e m a t i c a l l y
These mode l s are d i f f e r e n t i a t e d as follows:
M o del FA-1,
s a t u r a t e d model
Model FA-2,
deletion
of p a t h from SN to TB (^2 2 ^ *
Model FA-3,
deletion
of p a t h f r o m SN to TB (y2 2 ) an(^
from A to TB
(all po s s i b l e paths estimated);
P ath
(y 2 i ) •
In the a nalyses of these m o d e l s the p a t h from attitude to intention
(y- 1 ) was c o n s i s t e n t l y significant.
* 1 1
A l s o consistent w i t h expectations,
the p a t h f r o m intention to target b e h a v i o r
p r o b a b i l i t y level
targ e t b e h a v i o r
in its critical value.
(y2 i) w a s
was not significant.
{$ 21)
was beyo n d the
When a p a t h from attit u d e to
tested in Mod e l s FA-1 and F A - 2,
This
.0005
this
f i n d i n g supports the F i s h b e i n - A j z e n
parameter
mo d e l in
c o n t r a s t to B e n t l e r and S p e c k a r t ’s modification.
C o n t r a r y to the p r e d i c t i o n s of b o t h theoretical approaches,
p a t h f r o m s u b j e c t i v e n o rms to intention
nificant
level.
However,
the r e l a t i o n s h i p
jective n orms and attit u d e
to 5.278,
(y ^ )
all p s < .00006).
(^j)
the
did not achieve a s i g ­
(covariation)
was significant
between sub­
(t-values f r o m 5.199
29
Table 1
Structural Model Specifications
Fishbein-Ajzen Models
FA-1:
®21} Y i r
(Figure 3):
^ 2 1 ’ Y n > y 1 2 ’ Y 21
F A - 3:
3 21/ Y n
BS-3:
BS-4:
BS-5:
BS-6:
BS-7:
(Figure 4):
321 j Y ll^ y 12’ Y 13* Y 23
321' Y i r
y 12’ Y 13 j Y 23
321* Y i r
y 12’ Y 13 j Y 23
321>
y
11# y 12’ Y 13
321’ y n j Y 12
321’ Y n * y l2} Y 13* Y 21
321' Y i r
INT^ TB
> Y 12
Be n t l e r - S p e c k a r t M o dels
BS-2:
SN^
Y 1 2 * Y 2 1 9 Y 22
F A - 2:
BS-1:
k,
Y 12 j Y 21
A,, SN,
PB,
INT, TB
30
P rior b e h a v i o r d a t a
BS-7.
(PB) were included in Mod e l s BS-1 thro u g h
The p a r a m e t e r s of these mode l s are p r e s e n t e d sch e m a t i c a l l y in
Figure 4 and are d i f f e r e n t i a t e d
Model
BS-1,
s a t u r a t e d mo d e l
(see Table 1) as follows:
(all possi b l e paths estimated);
Mo d e l BS-2,
d e l e t i o n of p a t h from SN to TB
Model BS-3,
d e l e t i o n of paths
from SN and A to TB
Mo del BS-4,
d e l e t i o n of paths
from A, SN, and PB to TB
(Y 2 2 and Y 21^
’
^21 * Y 22* and Y 23^;
M o d e l BS-5,
d e l e t i o n of paths from A,
SN, and PB to TB;
and of p a t h from PB to INT (y 2 i> y 2 2 ’ y 23*
M o d e l B S - 6 , d e l e t i o n of paths
from SN and PB to TB
M odel BS-7,
from SN and PB to TB,
d e l e t i o n of paths
Y 13^ ’
(y 2 2
and
and Y 23^'
of
p a t h f r o m PB to INT (y 2 2 > Y 2 3 > and Y 1 3 ) •
T a b l e s C5 t h r o u g h Cll in A p p e n d i x C pres e n t the m e a s u r e m e n t and
causal m o d e l p a r a m e t e r estimates for these BS-series models.
to f a c i l i t a t e comparisons of the significance
e stim a t e s b e t w e e n models,
In order
levels of the p a r a m e t e r
t -values of es t i m a t e d paths and of the
e s t i m a t e d r e s i d u a l s of i n tention and target b e h a v i o r for all models
t e sted are p r e s e n t e d in T a b le 2.
Of the m o d e l s
in w h i c h a p a t h from attitude to target b e h a v i o r
(y 2 i) was tested,
a signi f i cant p a r a m e t e r es t i m a t e was a t t ained only
once and that was
in a m o d e l w h i c h ex c l u d e d all paths from prior
behavior
(Model BS-7).
This f i n ding supports the F i s h b e i n - A j z e n model
as does the f i n d i n g that the p a t h from attit u d e to intention
significant
(maximum £ -
Where paths
.0250)
(Y-q) was
for all models tested.
f rom p r i o r b e h a v i o r were estimated, p a r a m e t e r estimates
31
ri
CO
CQ
•3c
•3c
r—1
•
vD
•3c
03
O
•
CM
1
1
1
•3c
•3c
*d"
H
•
to
1
1
1
1
rr•
rH
•3c
He
O
CM
•
-d-
•3c
■3c
LO
CM
•
to
■3c
He
LO
LO
•
•d
■3c
•3c
d
CM
•
to
0
i-H
•H
cd
P
t
o
TJ
0
i-H
cd
P
i
o
p p
^ ^ ^—
VO
1
CO
CQ
LO
1
CO
CQ
cd
C/3
0
i-H
nS
>
cn
i-H
0
T3
O
S
-d"
1
CO
CQ
to
CQ
ii
X
cd
H
p
•3c
\D
•
CM
\0
rH
•
rH
1
1
•3c
He
■'dCM
•
LO
CM
1
CO
CQ
He
•3c
CM
00
»
CM
LO
oo
«
•3c
•3c
O
to
•
■<d-
r—1
1
CO
CQ
•3c
rCM
•
CM
LO
03
*
•3c
•3c
LO
CM
•
■<d-
g
0
0
■3c
•3c
00
03
•
CM
1
VO
00
•
O
CD
i-H
•3c
•3c
O
to
•
vO
rH
CM
•
rH
•3c
•3c
rH
oo
•
CM
to
1
co
CQ
0
•3c
'Cj'
to
•
CM
*
He
ro
•
LO
0
X ■
•3c
•3c
00
CM
•
'd-
LO
vD
•
I
3
1
1
1
1
3
3
1
1
1
1
1
1
1
1
1
1
1
1
He
•3c
O
O
•
*
He
00
•
'd-
•3c
•3c
00
■<d•
vD
•3c
•3c
O
o
•
to
■3c
•3c
LO
CM
•
to
He
•3c
OO
r•
vD
•3c
He
to
■'d•
d
He
He
■<d03
•
•3C
•3c
CM
O
•
LO
to
LO
•
rH
He
■3c
03
03
•
d
to
LO
•
rH
\D
•3c
■3c
LO
CM
«
\D
1
1
03
•
LO
vD
•
\D
rH
•
to
C/3
y
\D 00
03 LO
• «■
rH CM
A
A
0
o
C
cd
o
•H
<4H
•H
0
o
G
cd
a
*H
<4-4
*H
& &
•rH ~*H
tf)
tf)
t+H t+H
o o
rH
0
>
0
rH
rH
0
>
0
rH
LO rH
O O
• •
■3c -3c
■3c
tf)
G
O
C/3
•H
•3c
•3c
LO
LO
-
o
CM
•
rH
vD
•
LO
LO
LO
•
•3c
•3c
00
03
•
d
rCM
•
rH
■3c
* -3c
CM
LO
•
LO
•3c
He
LO
rH
•
LO
•3c
He
oo
to
•
to
•3c
•3c
LO
rH
•
LO
•3c
He
d
to
•
to
I-
0
*
G tf>
rH 0
cd P
>
cd
•H
p >
0
O
U
X
P
ccj
CO
1
<
PL,
Cl,
He
•3c
*d"
O
•
LO
rH
to
■
1
3
1
1
1
1
1
1
1
tf)
i-H
0
o
CM
1
<
PP
<
PL,
He
He
LO
o>
•
'Cj'
to
CM
»
3
1
i
■'d03
•
1
1
•3c
•3c
'd1
1
•
to
0 T3
C/3
0
3
rH
cd
>
i
P
tn
<D i”H
rG cd
p S
p
- o
vD C
to
CM X)
p
Cp cd
O T?
G
0 cd
N p
•H tf)
tf)
tf)
tH
1
<
PL.
X
p
ccj
CL,
He
•3c
•3c
•3c
roo
•
'Cj'
CM
rH
•
1
l
i
E2
T— 1 t— 1
rH -f
?*- <
E2
t-H
CM f
rH Z
CO
E-h
2
I-H
to +
rH CQ
CL,
to
to
•
rH
CQ
rH H
CM +
>- <
CM
rH
•
rH
1
CQ
H
CM f
CM 2
>CO
1
1
v£>
to
CQ
H
to f
CM CQ
>- Cl,
CQ
E+
rH E-1
CM Z
CO. I-H
C/3
0
O
G
cd
•H
Eh
cd
>
i— i
cd
3
TJ
•rH
c/3
0
OS
•3c
•3c
rrH
•
LO
He
•3c
03
to
•
to
E2
I-H
CQ
E-
0 cd
»-H
P h -G
g 0
cd P
tf) 0
Eh
0 P
X
P
P 0
P
o G
P -H
0
G
Q
cd
0
P
cd
32
w e r e c o n s i s t e n t l y s i g n i f i c a nt b o t h to intention
target b e h a v i o r
(£ <
.00006).
(jd <
.00006)
and to
These findings support one of Bentler
and S p e c k a r t ’s h y p o t h e s i z e d modifications.
N o p a t h was h y p o t h e s i z e d from subjective norms to target behavior,
ther e f o r e the f i n d i n g of n o n s i g n i f i c a n c e for this p a r a m e t e r estimate
^22^
saturatec* m odels was expected.
However,
co n t r a r y to e x p e c t a ­
tions h y p o t h e s i z e d b y b o t h the Bentle r - S p e c k a r t and F i s h bein-Ajzen
models,
the p a t h f r o m s u bjective norms to intentions
significance
(£ <
.05) o nly in Models BS-5 and BS-7.
w h i c h e x c l u d e d all p a t h s f r o m pr i o r behavior,
a model
c o n s i s t e n t w i t h F i s h b e i n and Ajzen's formulation.
path
(Y-j^p was n °t s i g n i f i c ant in Models FA-1,
(Y-j^
achieved
These were models
specification
Interestingly,
this
F A - 2, or F A - 3 w h i c h did
n o t include p r i o r b e h a v i o r data.
Model BS-5 was a f o r m u l a t i o n of the F i s h b e i n - A j z e n model w h i c h
i ncluded p r i o r b e h a v i o r d a t a in the m e a s u r e m e n t model only.
That is,
the p a t h s e s t i m a t e d were f r o m attitude and subjective norms to intention
only,
and just f r o m i n t e n t i on to target behavior.
causal p a t h s e s t i m a t e d were significant;
s t rengths o f the paths,
n o r m s to i n tention
In Model BS-5 all
but in looking at the r e l a t i v e
it is apparent that the p a t h from subjective
(Y-j^* t-value = -2.984, £
= .0028)
d i d not achieve
the level of s i g n i f i c a n c e r e a c h e d by the p a t h f r o m attitude to intention
(Yu*
t - v alue = 6.302, p < .00006)
ta r g e t b e h a v i o r
Model
or by the p a t h from intention to
t-va l ue = 6.480, £ < .00006)
(see Table C 8 ) .
BS-7 d i f f e r e d f r om Model BS-5 only in the a d d ition of a
p a t h f r o m a t t i t u d e to target be h a v i o r
and Speckart's
(1979)
(Y 2 j) anc^ was a test of Bentler
first model modification.
In this model also,
33
c o m p a r i s o n of t -values and the i r a s s o ciated p r o b a b i l i t i e s for the
estimated path parameters
(presented in Tables CIO and Cll)
the p a t h from subj e c t i v e n o rms to intention
£
=
.0366)
^ _ v a ^ue = "2.086,
did not a c h ieve a level of significance as h i g h as that
reac h e d b y the p a t h f r o m a t titude to intentions
£ <
.00006)
n o r m s and a t t i t u d e
However,
(^i^
.00006).
was d e m o n s t r a t e d in all m o d e l s
Also,
(f^P
1.770, £ =
il 2 1 * Ji”v a l u e =
a strong r e l ationship b e t w e e n subjective
(t-values >
in Model BS-7, when attitude was allowed a
d irect p a t h to t a rget b e h a v i o r
behavior
(Y-^* t> v a l u e = 6.409,
or the p a t h f r o m attitude to target b e h a v i o r
3.145, £ = .0016).
4.956, £ <
(Y 2 i)>
failed to r e a c h the
t^le P at^ fr o m intention to target
.05 level of signif i c a n c e
b e t w e e n subje c t i v e n orms and attitude
Qp <
.00006).
subje c t i v e norms
($ 3 ^) and
(^ 2 1 ^ was u n i f o r m l y h i g h in all
The r e l a t i o n s h i p b e t w e e n p r i o r b e h a v i o r and
(^3 2 ^ was
nonsignificant
(£ > .05).
E s timates o f the p a t h from intentions to target b e h a v i o r
w e r e e x t r e m e l y variable.
behavior data
sign i f i c a n t
(£ <
.002).
FA-2,
and FA-3),
However,
.00006)
behavior
this p a t h was co n s i s t e n t l y
w h e n a l t e rnate paths to target
this p a r a m e t e r dec r e a s e d
w h i c h included p r i o r b e h a v i o r d a t a
e s t i m a t e for this p a t h
(^2 1 ^
In those models w h i c h did not include p r i o r
(Models FA-1,
b e h a v i o r w e r e tested,
p <
(t>value =
.0768).
The r e l a t i o n s h i p b e t w e e n attitude and p r i o r b e h a v i o r
models
shows that
(f^)
in value.
In the models
(BS series m o d e l s ) , the p a r a m e t e r
r e a c h e d signif i c a n c e
(t-values > 4.00,
only in m o d e l s w h i c h d e l e t e d all oth e r p a t h s to target
(Models BS-4 and BS-5)
f r o m a t t i t u d e to target b e h a v i o r
and in Model B S -6 w h i c h included a pa t h
(Y 2 P *
Apparently,
the v a l u e of this
34
p a r a m e t e r is i n v e r s e l y r e l a t e d to the a v a i l a b i l i t y of alt e r n a t e paths
to target b e h a v i o r and the inclusion of p r i o r behavior.
w h e n o t her varia b l e s
are included in the model,
In other words,
intention is a less
influential d e t e r m i n a n t of target b e h a v i o r than the other variables.
In summary, w h e n the t-values of estimated p a t h p a r a meters are
compared,
the only p a t h s d e m o n s t r a t i n g consistent significance are
those from p r i o r b e h a v i o r to intention
Cy-^* £ <
from p r i o r
b e h a v i o r to t a rget b e h a v i o r
(Y 2 3 > £ <
intention
A l s o consistent was the finding that the
(y^j
£ <
.023).
*00006),
.00006),
p a t h f r o m s u b j e c t i v e n orms to target b e h a v i o r
(£ >
.05) w h e n e v e r it was tested.
and from attitude to
(Y2 2 ) was n o n s i g n i f i c a n t
All other paths v a r i e d in levels of
s i g n i f i c a n c e d e p e n d i n g on the inclusion or exc l u s i o n of other variables
in the model.
This
i n s t a b i l i t y suggests a n e e d for model re- s p e c i f i c a t i o n
to a t tain c o n s i s t e n c y of results.
I n s pection of the t-va lues for the residuals
of target b e h a v i o r
reve a l s that w h e n p r i o r b e h a v i o r is not p e r m i t t e d a direct path' to target
behavior
(Models BS-4,
BS-5,
B S - 6 , and BS-7),
this residual increases
f r o m a n o n s i g n i f i c a n t 1.27 or 1.53 to a significant level
(£s <
.003).
Th e a d d i t i o n of a p a t h f r o m attitude to target b e h a v i o r w i t h o u t a pa t h
f r o m p r i o r b e h a v i o r to targ et b e h a v i o r
result
(Models B S - 6 and BS-7)
in an i n s i g n i f i c a n t ^t-value for this residual.
did not
The t-values
the r esidual of inten t i o n w ere c o n s i s t e n t l y significant
(all £ s <
for
.003).
C h i - S q u a r e G o o d n e s s - o f - F i t Tests
The c h i - s q u a r e G o o d n e s s - o f - F i t test is a test of the m o d el's fit
to the o b s e r v e d data,
that is,
the v a r i a n c e u n a c c o u n t e d for b y the model.
Ta b l e 3 p r e s e n t s the c hi-squares,
degrees of freedom,
associated
35
oo
o
INrH
CM
■Tf
•
Tests
Goodness-of-Fit
•
rH
rH
o
O
o
rH
rH
V
V
04
o4
V
V
V
o
o
o
CM
CM
CN
■Tf-
LO
nO
rH
rH
rH
CN
T“ l
r—
^NrH
#
O
CN
LO
o
04
cm
rH
d"
03,
LO
•
rH
•
rH
V
Chi-Square
rCN
oo
to
\D
LO
CN
to
.
rH
CN
04
CQ
*\
E-<
LO
o
V
pj
°
|CN
•
d"
LO
o
LO
rH
00
OO
•
CN
LO
o
O
to
•
to
rH
o
o
o
03
\D
o
LO
03
•
CN
LO
CN
i-H
r—I
•
to
o
•
V
V
V
V
V
V
eg
Ph|
04
P-.I
Ph|
Ph|
to
00
to
nO
to
f'to
03
00
LO
to
nO
nO
oo
LO
*
LO
LO
rH
d"
NO
•
LO
LO
to
©
o
o•
•
to
oo
,o
E-
LO
•
rH
LO
d"
LO
•
rH
o
nO
to
nO
o•
*4-1 Id"
T3 [to
o>
CN
rH
.
CN
CN
rH
oo
oo
•
o
©
•
•
rH
NO
CN
d"
to
03
rH
d"
•
•
•
NO
NO
d"
o
rH
d"
oo
CN
•
NO
O
o
CN
rH
rH
CN
rH
2
I-H
CQ
H
*\
CQ
E-*
C l,
2
HH
•s
2
CO
•s
2
co
•s
CN
CN
>-
<
CNI
<C
t—
to
\
a»
u
3
00
•rH
Uh
/—\
■Tf
CN
CN
<D
P
0
oo
>-
*\
rH
CN
>
rH
CN
>-
'— '
CO
I— 1
<D
T3
O
S
o
<D
tsl
•i—>
<
CN
r“1
>-
CN
r-1
>-
>•s
r“1
CN
r“1
CN
CO.
ca
JO
X
CO
•H
CL,
CN
to
CN
CN
CN
>•s
to
>•S
to
to
>•s
to
rH
rH
>-
>•s
>to
rH
CD
-d
o
S
cd
ca
to
CN
•H
CO
rH
o
CD
O'
CNI
>-
•s
to
tL*
v
— '
u
*\
>•s
CN
rH
>-
n
CN
T— 1
>-
rH
CM
T“ l
>-
*\
>-
>-
•s
•s
rH
rH
CN
CO.
>-
rH
.
>-
>-
CM
CN
T“ l
>•S
>-
>-
>-
CN
rH
>-
rH
CN
>(N
rH
>-
*\
>-
>•s
rH
*\
rH
rH
rH
rH
CN
CO.
CN
CO.
CN
CO.
CN
CO.
CN
CO.
CN
CO.
rH
CN
to
d"
LO
co
CQ
C/3
CQ
C/3
CQ
co
CQ
CO
CQ
C/3
o
•H
<D
CN
rH
>-
>-
1
rH
1
<
P-,
CN
1
<
P-,
to
i
C
tu
O
CD
t-H
p
e
CD
CQ
•
i
1
NO
co
CQ
1
co
CQ
36
probabilities,
and ratios of chi-square d i v i d e d b y the degrees of
f r e e d o m for e a c h model.
The x 2 stat i s t i c p r o v i d e s a test of the p r o p o s e d model
a g a i n s t the general a l ternative that the MVs
variables)
(measured
are s i mply c o r r elated to an ar b i t r a r y extent.
If the x 2 is large c ompared to degrees
of freedom,
one
co ncludes that the m o d e l does n o t app r o p r i a t e l y m i r r o r
the causal p r o c e s s that generated the data.
(Bentler,
1980, p. 428)
A n o n s i g n i f i c a n t chi- s q u a r e value,
then,
supports the hypothesis
that
the m o d e l p r o v i d e s a p l a u s i b l e r e p r e s e n t a t i o n of the causal process,
that is,
the chi-s q u a r e
indicates w h e t h e r or n o t the factors specified
in the c o n f i r m a t o r y factor analysis extract sufficient va r i a n c e so that
the r e s i d u a l s are nonsigni f icant.
The ratio of c h i - s q u a r e d divided by
its d e g rees of f r e e d o m is also an index of g o o d n e s s - o f - f i t , w i t h a
b e t t e r fit b e i n g i n d i c a t e d b y a smaller ratio.
si g n i f i c a n c e test for this ratio,
There is n o asso c i a t e d
however.
A p r o b l e m that arises from the ex c l u s i v e use of the chi-square
G o o d n e s s - o f - F i t test for the e v a l u a t i o n of a model's fit to the d a t a is
that the c h i - s q u a r e v a r i a t e is a direct function of sample size and the
n u m b e r of p a r a m e t e r s
analysis
estimated.
Consequently,
is n o t a p p r o p r i a t e for small samples)
p a r a m e t e r s to b e estimated,
w i t h large samples
(this
and a large n u m b e r of
this statistical test w o u l d result in the
r e j e c t i o n o f v i r t u a l l y all models.
T h a t is to say that the appropriate
s t a t istical c o n c l u s i o n w ould be that the residual m a t r i x contains s i g ­
n i f i c a n t a d d itional
i n f o r m a t i o n that could be exp l a i n e d by a bett e r model.
37
W h i l e the v a l u e of x 2 depends on sample size,
the associated degrees
o f f r e e d o m are only d e t e r m i n e d b y the n u m b e r of var i a b l e s and
h y p o t h e s i z e d factors
in the model.
T he r e s ults r e p o r t e d in Table 3 show that the chi-squares for"'all
F A series m o d e l s a t t a i n e d a similar nons i g n i f i c a n t
(.20 < £ <
.10).
level of proba b i l i t y
The chi- s quares for all BS series models,
which
i ncluded more v a r i a b l e s along w i t h their a s s o c i a t e d para m e t e r s and
c o n s e q u e n t l y a d d itional m e a s u r e m e n t error, were at less than the
.005
level o f s i g n i f i c a n c e w h i l e the sample size was the same for b o t h series.
A c c o r d i n g to B e n t l e r and Bonett
(1980),
one m e t h o d of addressing
this p r o b l e m is to inspect the absolute values of r e s i duals which
p r o v i d e an e s t i m a t e of the amount of statistical
f r o m the data.
However,
infor m a t i o n extracted
(This i n f o r mation on residuals is p r e s e n t e d in Table 2.)
a "key ingr e d i e n t " in a p p ropriate statistical m e t h o d o l o g y for
comparisons of causal m o d e l s
in their view is the use of hierarchical
(nested) m o d e l s to p r o v i d e a chi-square diff e r e n c e test b e t w e e n models.
Th e p r i m a r y use of the i n f o rmation p r e s e n t e d in Table 3 is for the
c o m p u t a t i o n o f t h ese c h i - s q uare d i f f erence tests b e t w e e n h y p o t h e s i z e d
models.
The r e s u l t s o f these difference tests are p r e s e n t e d in Table 4.
C h i - S q u a r e D i f f e r e n c e Tests
The chi-s q u a r e d i f f e r e n c e test is based upon the rationale that the
d i f f e r e n c e b e t w e e n c h i - s q u a res is also d i s t r i b u t e d as chi-square with
degrees of f r e e d o m equal to the d i f f e r e n c e b e t w e e n the n u m b e r of p a r a ­
m e t e r s e s t i m a t e d b y each model.
This statistic is us e d to test the
importance o f the p a r a m e t e r s that d i f f e r e n t i a t e the mod e l s and to assess
the r e l a t i v e a d e q u a c y of the mod e l s
in e x p l a i n i n g the observed data.
38
T able 4
C h i - S q u a r e D i f f e rence Tests for Model Comparisons
Fishbein-Ajzen Models
xi
df
£
FA-1/FA-2
.7983
1
' .50 < £ <
.30
F A - 1/ F A - 3
1 .9527
2
.50 < £ <
.30
F A -2/FA-3
1 .1544
1
.30 < £ < .20
Bentler-Speckart Models
xi
df
£
BS-l/BS-2
1 .3830
1
.30 < £ < .20
BS-2/BS-3
.0582
1
.80 < £ < .70
BS-l/BS-3
1 .4412
2
.70 < £ < .50
BS-3/BS-4
50 .9726
1
£
< .001
B S-4/BS-5
2 0 .3198
1
£
< .001
BS-4/BS-6
.1995
1
.70 < £
BS-6/BS-7
13 .8699
1
. £
< .001
B S-3/BS-5
71 .2924
2
£
< .001
BS-3/BS-7
64 .6430
1
£
< .001
< .50
39
The null h y p o t h e s i s a p p r o p r i a t e for these tests is one of model
equivalence.
The c h i - s q u a r e differences,
associated probabilities
degrees of freedom,
and
for the model comparisons are re p o r t e d in
Table 4.
For the models w h i c h did n o t include p r i o r b e h a v i o r d a t a
series,
see T a b l e 4),
be rejected.
the null h y p othesis of model equivalence cannot
Essentially,
this means that the model with a p a t h to
t a rget b e h a v i o r f r o m i n t e n t i o n only
(Model F A - 3)
m o d e l s w i t h a d ditional p a ths to target b e h a v i o r
Thus,
(FA
w i t h p r i o r b e h a v i o r d a t a excluded,
is equivalent to
(Models FA-1 and FA-2).
the F i s h b e i n - A j z e n model
is
s u p p o r t e d as the m o s t p a r s i m o n i o u s and adequ a t e e x p l a n a t i o n of the
o b s e r v e d data.
When p r i o r b e h a v i o r d a t a are included
T a b l e 4),
for M o d e l s
deleted),
(BS series models,
see
the null h y p o t h e s i s of mo d e l e q u i valence cannot be re j e c t e d
BS-1
(the s a t u r a ted model),
and BS-3
BS-2
(path from SN to TB, y
(path f r om A to TB, y a l s o
deleted).
These results
support the m o d e l w i t h p a t h s to target b e h a v i o r from p r i o r b e h a v i o r and
intention, b u t n o t from a t t it u d e and subjective norms to target b e h a v i o r
as the best fit to the o b s e r v e d data.
Model
BS-2 w h i c h includes the at t i t u d e to target b e h a v i o r p a t h
S p e c k a r t ’s e x p a n d e d model)
model
C o m p a r i s o n of Model BS-3 w i t h
(Bentler-
does not support the B e n t l e r - S p e c k a r t
(1979)
in that the a ttitude to target b e h a v i o r p a t h does not fit the
o b s e r v e d d a t a s i g n i f i c a n t l y b e t t e r than the model w i t h this p a t h omitted.
/
C o m p a r i s o n of Model
BS-3 w i t h M o d e l BS-4
(paths to target b e h a v i o r
d e l e t e d from all v a r i a b l e s w i t h the e x c e p t i o n of intention,
in r e j e c t i o n of the null h y p o t h e s i s
of model equivalence.
results
This hypothesis
40
must also be r e j e c t e d f o r c omparisons of Model BS-3 w i t h subsequent
BS series mode l s as well,
indicating the n e c e s s i t y of paths from p r i o r
b e h a v i o r to b o t h i n t e n t i o n and to target b e h a v i o r
(y-j^ and Y 2 3 ) •
lower ratio for this m o d e l of y 2 to degrees of f r e e d o m
(1.5456),
comp a r e d to the v a l u e of that ratio for the other models,
The
as
supports the
c o n c l u s i o n that this m o d e l provi d e s the best fit to the ob s e r v e d data
of the m o d e l s
tested.
These findings support the B e n t l e r -Speckart
(1979) m o d i f i c a t i o n c a l l i n g for paths from p r i o r b e h a v i o r to both
i nten t i o n and target behavior.
B ecause of an equal n u m b e r of degrees o f freedom, Models BS-3 and
B S -6
(no p a t h from PB to TB, Y 2 3 > ^ ut
inclusion of a p a t h from A to TB,
Y 2 1 ) c a nnot be d i r e c t l y c o m pa r e d b y a y 2 d i f f erence test.
However,
c o m p a r i s o n o f the x 2 /df ratios of these models supports the b e t t e r fit
o f Model BS-3 to the data.
Discussion
Befo r e f u r t h e r d i s c u s s i o n of the results of this study,
is in order.
one caveat
It m u s t b e b orne in m i n d that this study c o n c erned only
a single b e h a v i o r a l content domain,
class attendance,
p o p u lation, u n i v e r s i t y s u m m e r school
o t h e r b e h a v i o r a l domains
students.
and a single
Replication,
and different popul a t i o n s
utilizing
of subjects,
is a
n e c e s s a r y p r e r e q u i s i t e to h y p o t h e s i z i n g changes in structural equation
models.
This p r e s e n t d i s c u s s i o n of results
direc t i o n s
domains,
for a d d i t i o n a l
is d i r e c t e d toward future
study in this area.
In c o n sidering behavioral
it s h ould be n o t e d that B e n t l e r and Speckart,
r e c e n t l y p u b l i s h e d s tudy in this a r e a
(1981),
in the i r most
found dif f e r e n t results
for d i f f e r e n t b e h a v i o r a l domains am o n g the same subjects.
41
To address the initial r e s e a r c h quest i o n p o s e d b y this study,
c o m p a r i s o n of the F i s h b e i n - A j z e n
(1979) modif i c a t i o n s ,
(1975) model with B e n t l e r - S p e c k a r t ’s
the results of this study lend support to the
B e n t l e r - S p e c k a r t h y p o t h e s i s of direct paths from p r i o r b e h a v i o r to both
i n t e n t i o n and target behavior.
However,
the results do not support
t h e i r model m o d i f i c a t i o n of a direct p a t h from attitude to target
behavior.
A n o t h e r area for a d d itional r e s e a r c h suggested by these findings
is a test of a r e s p e c i f i c a t i o n of the causal model wi t h subjective norms
and prior behavior acting upon
(causal to)
attitudes along with
h y p o t h e s i z e d direct effects of p r i o r b e h a v i o r on the other endogenous
variables
of inte n t i o n and target behavior.
The finding in this study
that the p a t h f r o m s u b j e c t i ve norms to intention
(y ^ )
achieved the
.05
level of s i g n i f i c a n c e in o n l y two of the mode l s tested was contrary to
b o t h the F i s h b e i n - A j z e n
p r e d i ctions.
(1975)
and B e n t l e r -Speckart
(1979) model
The F i s h b e i n and A j z e n formu l a t i o n does al l o w the weights
of a t t i t u d e and s u b j e c t i v e norms to v a r y w i t h the type of behavior,
,with the context or s i t u a t i o n in w h i c h the b e h a v i o r is to be performed,
w i t h the target,
and w i t h individual differences b e t w e e n actors.
w hile the h y p o t h e s i z e d p a t h
(causal relationship,
But,
was nonsignificant,
s u b j e c t i v e norms and a t t i t u d e exh i b i t e d a strong r e l a t i o n s h i p
(cor-
4
relation)
in all m o d e l s
(minimum t-value of cj^ of 4.956).
This
c o v a r i a t i o n s uggests the p o s s i b i l i t y of a r e s p e c i f i c a t i o n of the model
b a s e d u p o n the n o t i o n that social
influences are indirect determinants
of inten t i o n t h r o u g h t heir effect on attitudes.
Thus,
a r e a sonable
r e s p e c i f i c a t i o n of the m odel wo u l d be a test of subjective norms prior
42
to a t t i t u d e in the causal m odel
specifications.
The m o s t r e c e n t l y p u b l i s h e d Bentler-Speckart
(1981)
study found
support for a m o del w h i c h p l a c e d intention as an equal deter m i n a n t of
t a r g e t b e h a v i o r a long w i t h attitude,
subjective norms,
b e h a v i o r r a t h e r than as a m e d i a t i n g variable.
of the original m o del
and pr i o r
While this m o d i f i c a t i o n
is d i f f e r e n t f r o m the h y p o thesis sug g e s t e d above,
it too supports the n e e d for additional research on v a r i a n t s of the
F i s h b e i n - A j z e n model.
Likewise,
^31^
the r e l a t i o n s h i p b e t w e e n p r i o r b e h a v i o r and attitude
was e x t r e m e ly s t rong in all models w h i c h included p r i o r behavior.
A r e a s o n a b l e r e s p e c i f i c a t i o n w o u l d be the p l a c e m e n t o f p r i o r b e h a v i o r
causal to attit u d e s as well as to intention and target b e h a v i o r in a
m o d e l r e specification.
As m e n t i o n e d earlier,
p a t h p a r a m e t e r estimates,
behavior
the instability of some
n o t a b l y the p a t h from intention to target
seems to i n dicate the lack of sa t i s f a c t o r y model
s pecifications.
goodness-of-fit
In m o del BS-3 w h i c h de m o n s t r a t e d the best overall
(y2 /df),
^21^ WaS nons^Snificaivt
the t>value for the estim a t e o f this p a t h
•
A n o t h e r p o s s i b l e e x p l a n a t i o n of the n o n s i g n i f i c a n t
path might
subj e c t i v e n o r m
lie in the b e h a v i o r a l d o m a i n of the target behavior,
a t t e n d a n c e d u r i n g a f i v e - w e ek sum m e r session.
class
The d e v elopment of
group i d e n t i f i c a t i o n and group c o h e siveness w i t h classmates for s u c h a
brief period
social
is o b v i o u s l y minimal.
This
leaves as p r i m a r y sources of
influence s i g n i f i c a n t others n o t d i r e c t l y involved in the target
behavior
itself,
such as parents,
spouse,
c e i v a b l e that u n d e r such conditions,
and friends.
these social
It is c o n ­
influences m a y have
43
g e n e r a l i z e d to a b r o a d s p e c trum of attitudes toward academic be h a v i o r
in general.
A j z e n and F i s h b e i n
(1980) p r e s e n t e d consistent findings
de r ived from a v a r i e t y o f target b e h a viors w h i c h included v o t e r
behavior,
c o n s u m e r behavior,
Be n t l e r and S peckart
be h a v i o r a l domains,
(1981)
e.g.,
and family p l a n n i n g behavior.
found that in the same sample,
exercise,
studying,
and dating,
ferent r e s ults w i t h r e s p e c t to p a t h significance.
Again,
However,
three
led to d i f ­
additional
studies across b e h a v i o r a l domains are indicated.
A t h i r d p o s s i b l e e x p l a n a t i o n m i g h t reside in the samples used in
the p i l o t and p r i m a r y studies.
The "significant others" of the study
qu e s t i o n n a i r e were d e t e r m i n e d b y me a n s of a pi l o t study on a sample of
su m m e r school
stude n t s at the same u n i v e r s i t y who attended the session
i m m e d i a t e l y p r e c e d i n g the s ession attended b y the subjects of the
p r i m a r y study.
T h e s e two samples did not d i f f e r s i g n i f i c a n t l y on 13
out o f 15 d e m o g r a p h i c v a r i a b l e s
variables
(see Tab l e Cl
in A p p e n d i x C ) .
The two
in w h i c h signi f i c ant d i f ferences were obser v e d were y e a r of
g r a d u a t i o n and e c o n o m i c status as m e a s u r e d by pa r e n t ' s income.
The
p r i m a r y s t u d y sample was composed of students w h o r e p o r t e d an earlier
date of g r a d u a t i o n and a g r eater p r o p o r t i o n of whose p a r e n t s had lower
income
levels.
Possibly,
s tudents who are closer to g r a d uation p e r ­
ceive the i n s t r u m e n t a l i t y of class attendance in att a i n i n g the goal
of
g r a d u a t i o n d i f f e r e n t l y f r o m students for w h o m the e x p e c t a n c y of g r a d u ­
at ion is not so immediate.
For them,
the social
influence might be
ov e r r i d d e n b y the p e r c e i v e d i n s t r u m e n t a l i t y r e f l e c t e d in attitude w h i c h
was m e a s u r e d b y the c o n s e q u e n c e s of the b e h a v i o r m u l t i p l i e d b y an
e v a l u a t i o n of these consequences.
Also,
the diff e r e n c e in economic
44
status b e t w e e n samples m i g h t have led to the omiss i o n o f relevant
"sign i f i c a n t others" for c o n s i d e r a t i o n in the que s t i o n n a i r e u s e d in
the p r i m a r y study.
The q u e s t i o n o f g e n e r a l i z a b i l i t y requires a t t e n t i o n at this point,
for the v a lue of theo r y lies
in its generalizability.
If separate
t h e oretical formu l a t i o n s were r e q u i r e d for each beha v i o r a l dom a i n
investigated,
the u s e f u l n e s s
of such theory w o u l d be questionable.
It is the u l t i m a t e goal of t h e o r y - t e s t i n g not only to prov i d e empirical
support or disproof,
but to define the limits of generalizability.
The b e h a v i o r o f interest
a b e h a v i o r that
community.
in this study was c l a s s r o o m attendance,
in its own right is of p r o p e r conc e r n to the educational
A t t e n d a n c e b e h a v i o r is likewise o f p r i m e interest to
organizations
in the industrial,
governmental,
or service communities,
w here s u c cessful and e f f i c i e n t o p e r ation depends h e a v i l y u p o n the
p r e s e n c e of o r g a n i z a t i o n m e mbers or employees.
The p r a c t i c a l
s i g n i f i c ance of iden t i f i c a t i o n o f those v a r i a b l e s
w h i c h can u l t i m a t e l y affect this b e h a v i o r is obvious.
For example,
group norms were s i g n i f i c a n t p r e d i c t o r s of this behavior,
if
then o r g a n i ­
zational inter v e n t i o n s t a r g e t e d at the de v e l o p m e n t o f group cohesiveness
and d e s i r a b l e norms m i g h t be v e r y appropriate.
f ound to be m o s t significant,
If p r i o r b e h a v i o r were
as this study indicates,
such as n e w - e m p l o y e e supervision,
indoctrination,
interventions
and p r o m p t a t t e ntion
to the first indic a t i o n s of atte n d a n c e pr o b l e m s mi g h t forestall the
e s t a b l i s h m e n t of a p a t t e r n of p o o r attendance behavior.
A p o p u l a t i o n o f stude n ts w i t h class attendance as a target
b e h a v i o r is not the same as a p o p u l a t i o n of w o r kers w i t h w o r k attendance
45
as target behavior.
However,
the ba s i c interrelationships b e t w e e n v a r i ­
ables h y p o t h e s i z e d by F i s h b e i n and A j z e n have b e e n found to p o s sess
w i d e generalizability.
So it seems that an initial study ap p l y i n g this
model and v a r i a n t s of it to the target b e h a v i o r of attendance could
serve as a r e a s o n a b l e p r e l i m i n a r y step toward the m o r e general a p p l i c a ­
t i o n o f the model to this b e h avioral domain.
46
Reference Note
1.
Pedhazur,
E. J.
Personal communication*
De c e m b e r 8, 1980.
47
Re f e rences
Ajzen,
I.,
§ Fishbein, M.
A t t i t u d e - b e h a v i o r relations:
analysis and r e v i e w o f e mpirical research.
1977,
Ajzen,
84,
I.,
social b e h a v i o r .
U n d e r s t a n d i n g attitudes and pred i c t i n g
E n g l e w o o d Cliffs, N.J.:
S e lf-perception:
d i s s o n a n c e phenomena.
Bern, D. J.
modeling.
1980.
P sychol o g i c a l R e v i e w , 1967, _74, 183-200.
S e l f - p e r c e p t i o n theory.
P. M.
Prentice-Hall,
An a l t e rnative inte r p r e t a t i o n of cognitive
ex p e r imental social p s y c h o l o g y .
Bentler,
P s y c hological B u l l e t i n ,
888-918.
§ Fishbein, M.
Bern, D. J.
A theoretical
In L. Berkowitz
N e w York:
(Ed.), Advances
A c a d e m i c Press,
M u l t i v a r i a t e analysis w i t h latent variables:
in
1972.
Causal
A n n u a l R e v i e w of P s y c h o l o g y , 1980, _31, 419-456.
*
Bentler,
P. M . , § Bonett,
D. G.
S i g nificance tests and g o o d ness-of-fit
in the analysis of c o v a r i a n c e structures.
1980,
P s y chological B u i l e t i n ,
88, 588-606.
Bentler,
P. M.,
§ Speckart
G.
M o d e l s of a t t i t u d e - b e h a v i o r relations.
P s y chological R e v i e w , 1979, 86^, 452-464.
Bentler,
P. M.,
§ Speckart,
G.
Att i t u d e s "cause" behaviors:
structural e q u a t i o n analysis.
A
Journal of P e r s o n a l i t y and Social
P s y c h o l o g y , 1981, 40^, 226-238.
Billings,
R.
S.,
§ Wroten,
S. P.
o r g a n i z a t i o n a l psychology:
C r i ticisms and suggestions.
A p p l i e d P s y c h o l o g y , 1978, ^3,
Calder,
B. J.,
§ Ross, M.
General L e a r n i n g Press,
Use of p a t h analysis in industrial/
677-688.
A t t i t u d e s and b e h a v i o r .
1973.
Journal of
Morristown,
N.J.:
48
Calder,
B. J.,
Thibaut,
§ Ross, M.
J. T.
Attitudes:
Spence,
Theories and issues.
§ R. C. Carson
in social p s y c h o l o g y .
In J. W.
(Eds.), C o n t e m p o r a r y topics
Morristown, N.J.:
General Learn i n g Press,
1976.
Cook,
T. D.,
§ Campbell,
D. T.
issues for f i e l d s e t t i n g s .
P u b l i s h i n g Co.,
Eagly,
A. H.,
fi H immelfarb,
L.
Peterson,
Fishbein,
Fishbein,
Ra n d M c N a l l y College
S.
A t t i t u d e s and opinion's.
Heider,
M.
A t t i t u d e and the p r e d i c t i o n of behavior.
in a t t i t u d e t h e o r y and m e a s u r e m e n t .
M . , £ Ajzen,
I.
In M.
Fishbein
N e w York:
B e l i e f , a t t i t u d e , intention and b e h a v i o r :
Reading, Mass.:
A t t i t u d e s and c o g n itive organization.
K. G.,
structural
Addison-
H. C.
£ Sorbom,
D.
LISREL I V :
E s t i m a t i o n of linear
e q u a t i o n systems b y m a x i m u m likelihood m e t h o d s .
Inc.,
Chicago:
1978.
A t t i t u d e s are alive and well and ga i n f u l l y employed in
the sphere of action.
F. N.,
research.
Journal of
21_, 107-112.
N a t i o n a l Educa t i o n a l Resources,
Kerlinger,
Row,
1975.
P s y c h o l o g y , 1946,
Kelman,
111.:
1967.
F.
Joreskog,
Evanston,
1957.
A n i n t r o d u c t i o n to t h e o r y and r e s e a r c h .
Wesley,
Ann u a l R e v i e w
29, 517-554.
A theory of c o g n i t i v e d i s s o n a n c e .
(E d . ) , R e a d i n g s
Wiley,
Chicago:
1979.
of P s y c h o l o g y , 1978,
Festinger,
Q u a s i - e x p e r i m e n t a t i o n d e s i g n 5 anal ysis
A m e r i c a n P s y c h o l o g i s t , 1974,
§ Pedhazur,
N e w York:
Holt,
E. J.
2 9 , 310-324.
M u l t i p l e r e g r e s s i o n in behavioral
R i n e h a r t and Winston,
1973.
49
Li, C. C.
Path ana l y s i s :
Press,
A primer.
Pacific Grove,
Calif.:
1975.
Maruyama,
G.,
§ McGarvey,
B.
E v a l u a t i n g causal models:
A n application
of m a x i m u m - l i k e l i h o o d analysis of structural equations.
B u l l e t i n , 1980,
Namboodiri, N.
Osgood,
K., Carter,
L. F.,
$ Blalock,
H. M., Jr.
C. E.,
Rokeach, M.
Bass,
N e w York:
McGraw-
Urbana,
111.:
§ Tannenbaum,
P. H.
The m e a s u r e m e n t of
U n i v e r s i t y of Illinois Press,
B e l i e f s , a t t i t u d e s , and v a l u e s .
H. C.
§ Sons,
Triandis,
Suci, G. J.,
1957.
San Francisco:
Jossey-
1968.
Triandis,
A t t i t u d e and a t t itude c h a n g e .
N e w York:
John Wiley
1971.
H. C.
E. Howe,
(Vol.
A p p lied
1975.
meaning.
H.
Psychological
87_, 502-512.
m u l t i v a r i a t e analy s i s and experimental d e s i g n s .
Hill,
Boswood
27).
Values,
Jr.,
attitudes,
§ M. M. Page
Lincoln:
and interpersonal behavior.
In
(Eds.), N e b r a s k a Sym p o s i u m on M o t i v a t i o n
U n i v e r s i t y o f N e b r a s k a Press,
1980.
Appendix A
Q u e s t i o n n a i r e for Pilot Study
51
Informed C o n s e n t for P a r t i c i p a t i o n in a Re s e a r c h Project
Y o u r p a r t i c i p a t i o n in a r e s e a r c h study of student attitu des be i n g
c o n d u c t e d b y A r l e n e Fredr i cks of the UNO p s y c h o l o g y department is
requested.
P a r t i c i p a t i o n involves filling out a q u e s t i o n n a i r e duri n g a
class m e e t i n g on the subject of some of y o u r attitudes and p r o v i d i n g
some p e r s o n a l
i n f o r m a t i o n about yourself.
All
i n f o r m a t i o n will b e confidential and the a n o n y mity of your
i
re s p o n s e s will b e guaranteed.
Y o u r responses will n o t be identified to
your
i n s t r u c tor/professor.
Yo u r d e c i s i o n on w h e t h e r or n o t to p a r t i c i p a t e in this study or to
w i t h d r a w f r o m the study at any time will
in no w a y p r e j u d i c e your
r e l a t i o n s h i p w i t h the i n s tructor or the university.
Y o u r s i g n a t u r e on this consent f o r m indicates y o u r w i l l i n g n e s s to
participate
in this s tudy and a u t horizes the use o f the information
c o l l e c t e d a l o n g w i t h c l a s s r o o m da t a for r e s e a r c h purpo s e s only.
There
are n o h i d d e n cond i t i o n s or m a n i p u l a t i o n s involved and y o u are free to
w i t h d r a w f r o m the s tudy at any time.
I will be h a p p y to a n s w e r any
q u e s t i o n s y o u m i g h t h a v e on this p r o j e c t and appr e c i a t e your cooperation.
A r l e n e Fredricks,
554-2704 or
334-1177
S i g n a t u r e__________________________________________________________________________
D ate
Investigator
Social S e c u r i t y N u m b e r __________________
Sex _________________
Marital
Status:
Age
Single
Married
Separated
Di v o r c e d
Widowed
___
___
___
___
Student S t a t u s :
N u m b e r of credit h ours y o u are cu r r e n t l y taking.
____
N u m b e r of cred i t h ours y o u expect to take in the fall.
E x p e c t e d date o f g r a d u ation
(if y o u are in a degree p r o g r a m ) .
C h e c k here if y o u are not in a degree program.
______
E m p l o y m e n t Status:
A v e r a g e n u m b e r of h ours that y o u wo r k p e r w e e k currently.
__
A v e r a g e n u m b e r of h o u r s that y o u expect to w o r k p e r week
this fall.
Race:
Citizenship:
Caucasian
Black
S p a nish
A m e r i c a n Indian
Asian
U.S.
Ot h e r
P a r e n t s ’ A p p r o x i m a t e A n n u a l Income (please check y o u r best estimate)
D o n ’t k n o w ________________________
Less than $10,000
___
B e t w e e n $ 1 0 ,000 and $14,999 ___
B e t w e e n $ 1 5 , 0 0 0 and $ 2 5,000 ___
O ver $ 2 5 ,000
___
E x p e c t e d grade in this class:
A
GPA:
>0
4.00
3.90- 3 . 9 9
3.75-3.89
3.50-3.74
3.00-3.49
2 . 50-2.99
2 . 00-2.49
1.50 - 1 . 9 9
Less than 1.50
, F
53
B r i e f l y list the b e l i e f s that come to mind w h e n y o u are as k e d the
f o l l o w i n g questions.
If y o u run out of space for y o u r answers, use the
b a c k of the sheet and indic ate the question number.
1.
W h a t d o y o u b e l i e v e a r e the advantages of a t t e n d i n g y o u r summer
school class(es) e v e r y day for the summer session?
2.
What do y o u b e l i e v e are the d i s advantages of yo u r a t t e n d i n g your
s umm e r school class(es) every day for the summer session?
3.
What do y o u b e l i e v e are the advantages of your b e i n g absent from
y o u r summer school class(es) once?
Several times?
One day a week?
Twice a week?
M o r e than twice a week?
54
4.
W h a t do y o u b e l i e v e are the d i s a d v a n t a g e s of your b e i n g absent from
y o u r s u mmer school class(es) once?
Several times?
One day a week?
T w i c e a w eek?
M o r e than twice a week?
5.
What else do y o u a s s o c i ate w i t h y o u r attendance and/or absence from
this summ e r school class?
6.
W h o are the indiv i d u a l s or groups of peop l e whose opinions or
influ e n c e is i mportant to y o u w i t h respect to y o u r atte n d a n c e and/or
a b s e n c e f r o m y o u r s u m m e r school class(es)?
Appendix B
56
I nformed C o n s e n t for P a r t i c i p a t i o n in a R e s e a r c h Project
Y o u r p a r t i c i p a t i o n in a r e s e a r c h study of student attitudes be i n g
co n d u c t e d b y A r l e n e F r e d r i c ks of the UNO P s y c h o l o g y Department is
requested.
P a r t i c i p a t i o n involves f i l ling out a q u e s t i o n n a i r e either
d u r i n g a class m e e t i n g or at home on the subject of some of your
attitudes, beliefs, and intentions about a t t e n d i n g or not a t t ending
c lasses and also in p r o v i d i n g some personal infor m a t i o n about yourself.
All i n f o r m a t i o n will b e confidential a n d the a n o n y m i t y of your
r e s p o n s e s will be guaranteed.
Y o u r r e s p onses will N O T be identified to
y o u r i n s t r u c tor/professor.
They will be us e d for r e s e a r c h p u r poses
only.
Y o u r d e c i s i o n on w h e t h e r or not to p a r t i c i p a t e in this study or to
w i t h d r a w from the s t udy at any time will in no w a y p r e j u d i c e y o u r r e l a ­
ti o n s h i p w i t h the i n s t r u c t o r or the university.
Yo u r signa t u r e on this consent f o r m indicates y o u r willi n g n e s s to
p a r t i c i p a t e in this study and authorizes the use of the information
co l l e c t e d a l o n g w i t h c l a s s r o o m da t a for r e s e a r c h pu r p o s e s only.
Y o u are
free to w i t h d r a w f r o m the s tudy at any time.
Y o u are also free to omit
a n y q u e s t i o n s that y o u do n o t desire to answer, but it w o u l d be a p p r e ­
c i a t e d if y o u w o u l d a n s w e r all questions on the in f o r m a t i o n sheet and on
the a ttitude q u e s tionnaire.
A f t e r c o l l e c t i o n of all the data, the entire study and h o w the data
that y o u h ave p r o v i d e d is to be interpreted and u s e d will be explained
to you.
I will b e h a p p y to a n s w e r any questions that y o u m i g h t have on this
p r o j e c t and a p p r e c i a t e y o u r cooperation.
A r l e n e Fredricks,
P s y c h o l o g y Dept.,
Signature
Date
I n v e s tigator
554-2704 or
(home) 334-1177
554-2592
57
INST R U C T I O N S FOR Q U E S T I O N N A I R E A D M I N I S T R A T I O N
Y o u are b e i n g a s k e d to p a r t i c i p a t e in a re s e a r c h study wh i c h involves
s t u d e n t attit u d e s and opinions about att e n d i n g or not a t t e n d i n g class.
As the Informed C o n s e n t F o r m indicates, p a r t i c i p a t i o n involves p r o v i d i n g
some i n f o r m a t i o n about y o urself, fill i n g out a q u e s t i o n n a i r e about y o u r
b e l i e f s and intentions c o n c e r n i n g a t t ending or not a t t e n d i n g class, and
a u t h o r i z i n g the use of c l a s s r o o m da t a about yourself.
If there are any
questions that y o u do not w i s h to answer, you may leave them blank.
However, it is important to the study to collect as m u c h of this
r e q u e s t e d i n f o r m a t i o n as p o s s i b l e and it w o u l d be v e r y m u c h appre c i a t e d
if y o u did a n s w e r all the questions.
It is v i t a l to the s t udy that yo u r responses be iden t i f i e d b y y o u r
Social S e c u r i t y Number.
However, y o u r responses will N O T be identified
to your i n s t r u c t o r and will be u s e d for the r e s e a r c h pu r p o s e s of this
st u d y ONLY.
A f t e r all the d a t a h a v e b e e n collected, y o u will be informed of
the entire scope of the s t udy and the m e t hods used and of h o w the data
that y o u h a v e s u p p l i e d will be inter p r e t e d and used.
I will be h a p p y to a n sw e r any questions that y o u mi g h t have on
this study.
M y p h o n e n u m b e r is on the Informed Consent F o r m (334-1177)
or y o u m a y contact me t h r o u g h the U N O Psy c h o l o g y Department at 554-2592.
T h a n k y o u v e r y much.
Ar l e n e Fredricks
PLEA S E NOTE:
IF A N Y OF THE Q U E S T I O N S A B O U T OTHER PE O P L E ' S O P I N I O N S OR
Y O U R D E SIRE TO C O M P L Y W I T H THE O P I N I O N S OF THESE OTHERS
(SUCH A S HUSBAND, WIFE, OR FIANCE) DO N O T APP L Y TO YOU,
PLEASE L E AVE THE R A T I N G SCALE BLA N K FOR T H A T Q U E S T I O N AND
W R I T E N / A (NOT APPLICABLE) A F T E R THE QUESTION.
58
i-i
s
O
••
p
0
p:
p
•H
0
P
••
X
r—H
P
p:
•H
O h co
o o 0
o
CD p o3
—t
CO 0 i
3 o PH
X
CD
P
Pi co 0
o3 CD >
0
B
• H co
CO P
e O 0
o CO p :
• H CD p
P od
n
co
CD p 0
—t
CO i
cr CD o3
Pi O
bO
co
P p
• H o3 o3
p p:
•H p p :
o3
o
€ o 0
CD o CO
P o3
t-H P
CD Ph o
PS
0
H
p :
p
p
•H
bO
•H
r—H
CO
•«
5
•N
od
o
o
bO
X
t-H
0
B
0
p
p
X
0
CO
••
p
0
p
p
:
•H
0
e
••
X
t-H
4->
p
bO
•H
t-H
CO
•H
o3
p:
o3
..
0
••
CO
•H
3
cr
X
o3
t-H
.O
o3
B
O
0
B
0
P
P
X
0
e
•H
p
0
p :
p
o3
0
S
0
S
Td
o
o
toO
X
i-H
P
p
bO
•rH
i—1
CO
•N
od
'o3
Pi
p
o
p
od
o
o
too
••
p
0
X
•H
P ••
0 CO
•H
p
0
o3 03
0 p:
£ o3
g
X
f—i
0
B
0 <5 X 0
P
p:
p p
P
•H
X
0
Pi
e P
• H 0 od
p : p : o
p
o
p
o3 b O
3 0
O £
X
0
<p
HH S
••
is
i-H
P
p
0
3
CO
bO
•H
•H
i-H
co
o3
p
o3
..
0
P
•H
c
•H
3
cr
p .•
0 CO
•rH
p
•.
X
H
0
B
0
P
P
X
0
o3 03
p
o3
g
0
£
0
p
<5
p p
•H
pi
e p
•H 0 T d
o
p p
p
p
•rH
c
p
p
•H
p
..
p
0
p
0
p
•H
3
cr
C
p
.•
..
p
g
o
<§
P
•H
c
p
X
t
—H
p
bO
•H
i—1
CO
p
•rH
0
.•
3
O
x
p
0
p:
0
p
3
cr
3
O
X
p
o
o3 b O
3 0
O S
X
0
<p
l— l £
o3 03 CO
•H
P
CO
o3 0 03
i-H p :
b O o3 o3
C O g
• H CO o
CO
3
0
Pi
O
0 P
co • H
0
p: P
0
p :
co
p
• o3
i—H
nj CO 0
p 2:
t—H • H
t-H o3 0
•H Ph
f
2:
0
3 >
O •H
X P
o
0
•r— k
Td
03
extremely
••
quite
0
p
3
cr
•H
•.
••
slightly
X
i—t
p
rP
toO
•P
i—1
CO
..
B
0
p
p
X
0
X
t-H
0
B
0
p
p
X
0
t-H
C
3
••
neither
••
0
Pi
•H
••
X
3
*-i
C
slightly
0
p
3
cr
•H
X
t
—H
T3
o3
Pi
X
i-H
0
t—H
••
•rH
quite
•«
X
t—l
0
S
0
p
p
X
0
Td
o3
Pi
0 nd
> 0
P co
o3 o3
3
t-H 0
o3 U
> 0
0 2:
P 3 CO
0 O
p X o
i-H
p
o 4h t-H
•H o
CO
Uh
o3 •s
0 CO
i-H i-H o3
i-H P h
0
P
5 i o3
X 0
CO 0 P h
o3
u §•
co o
P P h Td
t-H
p
•H
3
o • O
Ph X
nd 03
C S P
•H
0
0
»N
X
i-H
0
0 CO i-H
o3
Pi
•H 0 O
t-H p : CO
e p
o3
3
i p
X •H p :
i-H
o
0 od 3
Pi 0 CO
•H P
i-H 0 P
P o
p : Ph
p
P
•H 0 •
P
S
X
P f“H
0 •H 3
t-H
•“i
o3 0
O Pi P
•H
CO
O
too P P
O
e
•H 0 p :
P P
•H Pi
P
o
:
p
CO
•rH
••
o3
:
o3
g
extremely
nd
o3
Pi
••
co
5
O
I— 1
rH
o
<p
to
03
Pi
P
03
B
P
3
O
X
0
o
o3
i-H
PH
od
t-H
3
O
rate
p •H
o3
e P
CD
pi X
o P
0 o3
PS CD
O 5
etf CD
Pi
0 E-*
Pi
o3
B CD
P
Q o3
P P
0 O
P P
o3
od
3 CD
O Pi
X co ##
o3
•
</)
10 CD
0 P o
O CD i-H
o3 £ i-H
i—t
o
p , 3 <p
O
e X CO
<u
o3
> m
0 • H od
CO
0
•N p
p j CD 0
p t-H P
•H PH O h
P
5
0
CO X P
CD CD P
i— 1
•H
o3 P
CJ O 0
co tin Pi
too *rd
P . t-H
•H P £
P O O
o3 • H rC
P P CO
i—i
i—i
o
<p
CO
o3
Pi
P
o3
C
P
3
O
X
Cl)
o
03
i—t
Ph
Td
t-H
3
O
to
P
Omaha"
The first page of the Questionnaire asks
income asks for your best estimate or, if you
have been out of school for some time and are
you, you may change it to "My annual income"
for personal information. The
have no idea, there is a place
now returning and consider that
and answer accordingly.
question on parents' annu;
for "Don't know."
If you
question irrelevant for
o3
p
Q
P
•H
P
rC
0
P
o3
0
S
0
p
H
:
X
rH
CD
•H
t—H
If you think it is extremely likely that the weather in Omaha is hot
scale as follows:
The weather in Omaha is hot in July.
likely
X
:
:
:
:
:
extremely
quite
slightly
neither
slightly
quite
July,
you
would
mark
:
unlikely
extremely
in
the
59
Social
Security Number
Sex _________________
M a r i t a l Status:
Age
Single
Married
Separated
Di v o r c e d
Widowed
___
___
___
___
Student S t a t u s :
N u m b e r of credit h ours y o u are cur r e n t l y taking.
N u m b e r of credit h ours y o u expect to take in the fall.
Ex p e c t e d date of g r a d u a t i o n
______
(if y o u are in a degree program).
C h e c k h e r e if y o u are not in a degree program.
______
E m p l o y m e n t Status:
A v e r a g e n u m b e r of h o u r s that y o u work p e r week currently.
A v e r a g e n u m b e r o f h o u r s that y o u expect to work p e r w e e k
this fall.
Race:
Citizenship:
Caucasian
Black
Spanish
A m e r i c a n Indian
Asian
U.S.
Other
P a r e n t s 1 A p p r o x i m a t e A n nual Income (please check y o u r best estimate)
D o n ’t k n o w
___
Less than $ 1 0 , 0 0 0 ________________
Betw e e n $ 1 0 ,000 and $ 1 4,999 ___
B e t w e e n $ 1 5 , 0 0 0 and $25,000 ___
O ver $25,000
___
E x p e c t e d grade in this class:
A ____, B
GPA:
, C
4.00
3.90-3.99
3. 7 5 - 3 . 8 9
3.50-3.74
3. 0 0 - 3 . 4 9
2. 5 0 - 2 . 9 9
2.00- 2 . 4 9
1 . 50-1.99
Less than 1.50
,D
, E
, F
unpleasant
61
%
62
e
o
TJ
0
p
o
3
3
•H
P
0
0
p
ps
3
subject.
x
rH
0
e
0
p
P
X
the
0
more
about
0
P
•H
3
cr
p
32
learn
w>
•H
fp
me
to
to
•i-t
r— l
0
3
3
3
o
enable
will
•H
0
e
T3
p
0
day
every
class
0
s
0
p
p
X
0
0
PS
3
e
p
o
m
3
•H
cr
0
32
0
0
p
•H
3
pH
cr
0
0
bO
3
•H
X
T3
3
p
•H
3
cr
4-1
p
w>
•H
P
o
x
0
0
0
p
•H
bO
3
T3
3
cr
p
x
bo
•H
p
o
t
-H
o
&
•H
C/3
in
•H
T3
3
•H
J
32
bO
0
p
P
to
0
C/3
32
p
0
3
32
p
•H
o
p
0
3
0
t/3
•H
3
oj
0
3
3
bO
•H
O
X
erf
TJ
3
3
0
P
•H
3
cr
rH
o
C/3
t/3
••
to
X
£
o
p
4-1
C/3
33
3
rH
bO
0
PS
X
•H
ip
3
o
t/3
•H
32
X
x
•H
ip
C/3
0
P
•H
3
cr
X
P
0
>
0
••
3
X
0
O
0
o
0
tuO
3
•H
T3
3
0
X
rH
0
PS
•H
ip
P
P
33
3
X
X
s
#
3
0
3
33
3
bO
3
•H
0
X
rH
0
PS
•H
ip
E
0
P
p
X
0
3
•H
32
P
W>
3
•H
T3
3
0
p
p
3
x
w>
O
3
•H
I
£
3
cr
X
rH
0
PS
•H
ip
0
p
6 E
32
32
•H
•H
■£
0
P
0
3
E
32
P
-H
£
0
3 32
3
3 O
P
32
W>
0
p
3
r
•H
•
*
0
• H #H
32 P
P
X
rH
0
PS
•H
rH
P
0
C2
X
X
x-S
s
s
S
03
O
t
0
t/3 3
t/3 •H
3
rH
o
t/3
•H
0
P
3
O
P
X
P
•H
3
cr
P
•H
3
E 33
••
X
rH
0
P
P
p
0
O
3
bO
•H
-H
in
E
0
X
*\
P
p
P
32
o
p 0
4-1 33
P
b/3 pi
P
X
0
0
3 £
3
C/3
0
T3
•H
>
O
32 Ip
rH
P
0
P
-p P
£
P
Ph
••
TJ
X
P
0
>
0
O
3
E
0 T3
#
bO
•H
0
p
3
•H
33
3
0
32
0
3 P
3)
00
0
32
T3
0
T3
P
3 0
3 ■H
O P
4H
X
3
o
P
x
0
32
0
P
p
0
4-1
3
rH
•H
32
cr
P
3
rH
E
0
P
p
X
0
3
t/3
rH
P
4-1
0
p
•H
3
C/3
£
E
0
P
p
X
0
0
^H
p
rH
t/3
0
T3
bO
t/3
t/3
••
P
•H
P
•H
>
•H
P
O
0 3
£ 3
*3
0
P
•H
3
cr
0
x
0
3
O *4-1
P
Sf
3
3
3
^H
p
p
•rH *H
X
bO
•H
rH
in
3
oj
O
•H
33
0
3 X
0
C/3
p
C/3
o
P
0
P
X
t
-H
rH
0
p
p
X
0
3
0
••
•H
rH
e
P
0
P
•H
3
cr
X
p
0
>
0
p
32
tuO
X
o
T3
O
rH
C/3
C/3
>?
0
e
rH
x
oj
3
e
0
W>
O
3
3
0
•H
0
P
rH
e
P
P
0
e
0
C/3
0
t/3
Oj
•H
X
o
3
0
0
4
1
PS
0
P
0
e
C/3
X
§
0
3
O
P
o
3
O
32
P
3
^H
C/3
p
p
oj
C/3
3
vD
0
W>
•H
p
oj
P
s
e
o
X
0
p
03
3
•H
0
x
0
P
P
3
oj
X
rH
p
x
x
P
P
p
p
p
O
•H
•H
32
e
o
0
p
3
3
3
0
e
0
C/3
0
p
•H
3
cr
3
3
X
PS
0
P
£
W>
0
0
0
0
32
O
•H
rP
T3
fl)
•H
P
0
PS
bO
p
O
p
Cl)
PS
3)
0
p
03
P
this
attending
cl)
PS
•H
ip
S
C
l3
PS
•H
to
My
X
e
p
O
■*H
P
3
E
0
P
P
3
0
t/3
32
3
O
3
O
•H
bO C/3
3 3
•H 0
0P
32
T3
£
X
4-1
3
3
0
X
^H
0
PS
•H
ip
63
P
X
p
o
o
p
c
o
•3
§
3
a
p
to
£
p
1
p
£
p
p
tuition.
X
CD
I— t
p
P
CD
•rH
for
3
X
P
£
p
•rH
to
a"
p
p
X
ccj
X
X
TJ
P
cr
paid
•rH
money
XbO
•P
3
bO
•rH
i-H
CO
•<
the
of
a waste
P
P
X
rH
■M
X
i-H
to
••
P
P
•rH
P
this
class
is
b0
•rH
cr
•.
from
X
i-H
P
absent
being
My
X
i-H
P
X
•H
i-H
p
•rH
O
o
bO
P
P
to
Ph
O
P
cr
P
o
PH
••
X
p
p
p
•rH
>
3
cr
to
to
ccj
•.
i-H
X
r-H
a
P
to
•rH
P
£
P
.
P
p
£
P
b£> P
P P
•rH to
*3 P
i
p
p P
p i-H
p P
ccj to
X
1 ?
to
•rH
P
P
£
P
P
•rH
£
X
P
,P
bO
•rH
P
ccj
i-H
to
sf £
••
P
X ccj
P tP
P P
>
P Ph.
P
P
•rH
3
o
to
P
cr
•1
i-H o
X
a •rH
P
to ccj
•rH P
X ccj
p i-H
i-H
P
£
P
P
p
X
P
Ph
bO X
P P
X
i-H
P
Pi
•rH
"d
P p
P ccj
i-H
to
CtJ p
to
X p
p
S
•rH
P
P
P
P
P
to
p
P
X
i-H
p
tX
•rH
i-H
x
o
Ph
p
p
p
t
P
p
•rH
P
P
rH
■M
o
p
X
••
ccj
P
p
•rH
X
CCj
bO
•rH
i-H
p
p
X
O
P
to
to
ccj
bO
o
"d
p
p
X
•rH
"d
•rH
>
X
rH
+->
o
X
p
Ph
X
ccj
"d
bO
•rH
i-H
to
••
X
tp
a
to
•H
•rH
P
ccj
3
H
aj
>
p
u
3
••
X
to
£
•rH
P
p
P
p
X
b£>
. P
•rH
"d
P
P
P
P
X
H
p
rX
•H
i-H
s
LO
X
i-H
P
P
P
,P
P
P
bO
•rH
3
i-H
£
to
P
3
P
•rH
to
X
to
3
cr
tX
to
ccj
to
p
o
•rH
P
to
P
3
p
cr
m
o
X
rH
o
£
rQ
P
P
P
O
£
P
P
E-
X
P
p
p
p
X
X
p
•rH
rH
aj
•H
p
p
p
ccj
£
p
p
d
p
u
p
rH
rd
P
p
O
•rH
P
ccj
£
p
P
ccj
P
•H
P
P
p
p
to
P
p
Mh
P
3
bO
P
•rH
P
l-H
X
o
X
P
X
to
•rH
ccj
•r->
3
X
t+n
O
P
i-H
P
a
P
p
o
r-H
P
P
X
o
p
<p
to
to
ccj
P
•rH
p
P
3
P
X
X
P
•rH
cr
P
O
P
>
i-H
to
P
P
to
to
ccj
o
P
o
to
to
i-H
£
o
.p
P
cr
bO
P
•rH
to
CtJ
•8 •rH
P
to
P
x
.1
P
i-H
X
CtJ
TJ
to
•rH
P
r-H
bO
§
p
•rH
cr
bO
p
P
bO
•rH
p
p
X
p
P
•rH
o
p
iH rH
X
TJ
to
P
X
t-H
P
P
•rH
to
O
lH
P
>
P
P
p
X
P
X
TJ
£
P
P
P
P
P
p
•rH
P
5
P
P
P
X
rP
•rH
P
P
P
P
•P
•rH
o
12.
to
o
pC5
*3
CCj
Pi
P
ccj
P
P
CCj
Pi
P
P
ccj
X
to
i-H
p
P
P
•rH
•rH
TJ
^3
p
p
(I.)
rX
P
X
o
0
•rH
i—t
>
Ph
•rH
P
p
to
extremely
r—I
p
•rH
•rH
■X
p
p
p
quite
iH
p
slightly
p
M
•rH
i—t
p
p
p
P
neither
P
.X
•H
slightly
ccj
quite
p
extremely
"d
p
"d
O
o
bO
P
P
ccj
p
p
o
Ph
£
•rH
bO
P
•rH
CO
to
•p
rJ
CN
T3
O
O
bO
64
m
P
P
O
•H
P
p
x
P
43
X
i-H
o
E
a>
p
p
X
o
X
r-H
o
E
a>
p
p
X
o
o
+->
•H
p
cr
o
+->
•rH
P
cr
••
X
i-H
P
43
bO
•H
r-H
tn
X
i"H
+->
4P
bO
•H
i-H
m
••
••
p
o
43
+->
•H
<D
P
X
i-H
P
43
bO
•rH
i-H
in
••
•
w
•H
Ph
o
O
r-H
m
P
O
+->
O
•
r-H
cti
ttf)
P
•rH
Ol
tn
•H
o
p
•H
P
cr
X
r-H
o
e
o
p
+->
X
o
P
O
O
ttf)
e
P
JO
P
43
P
O
43
+->
•rH
a>
P
••
m
•rH
••
w
X
m
r-H
cti
P
i-H
43
(J
bO
•H
t/)
i-H
•H
m
43
+->
P
•rH
o
+->
•rH
bo
P
P
•H ,
cr
X
P
P
+->
X
i"H
w
o
P
E
•rH
CD
P
+->
P
X
P
•rH
o
43
o
p
43
O
O
bO
bO
p
•H
i-H
rH
cti
43
+->
X
f-H
o
E
o
P
+->
X
O
••
m
•H
O
m
p
P
o
o
m
•rH
rP
P
•rH
<D
P
cti
P
bO
43
bO
•rH
43
ctJ
bfi
£
•rH
4->
+->
O
bo
P
P
cti
in
m
r-H
o
01
•rH
43
+->
P
•H
r-H
i-H
o
*:
bO
P
•rH
s
Ph
o
a
.
to
•
'3'
LO
•
O
+->
•rH
P
cr
••
X
f-H
+->
43
bO
•rH
r-H
in
•*
P
O
43
+->
•rH
O
P
•'
X
r-H
+->
43
bO
•rH
r-H
to
••
o
+->
•H
P
cr
♦•
rX
-H
a>
E
o
P
P
X
o
H3
O
O
bO
•H
5
o
43
O
P
o
s
•H
+->
P
P
ttf)
P
•rH
r-i
o
cti
43
••
o
E
•H
P
P
O
•H
P
cti
O
P
(j
<D
W
cti
43
O
P
m
in
o
•rH
P
•H
>
•rH
P
o
ct*
P ••
o w
•rH
P
P X
O 1-H
•H
bO B
p c3
•H p
tn
in P
•h c
.
\D
X
r-H
o
E
o
p
p
X
o
••
••
X
i"H
p
43
bO
•rH
r-H
w
P
o
43
P
•rH
o
p
••
•
•
P
cti
43
cti
4=1
o
p
•rH
P
cr
P
o
uh
P
P
X
r-H
p
43
ttf)
•rH
r-H
w
o
p
•H
P
cr
••
•rH
o
p
•rH
P
P
o
p
X
i-H
•rH
rd
O
43
4_>
q_l
o
P
o
•rH
in
P
o
••
••
nd
P
rt
••
X
i"H
o
E
o
p
p
X
O
P
o
o
bo
o
p
•H
p
cr
X
i"H
p
43
bO
•rH
r-H
m
X
p
43
bO
•rH
i-H
in
••
P
o
43
P
•rH
a>
P
••
X
i-H
P
43
bO
•rH
i-H
w
••
a>
p
•rH
P
cr
••
X
r-H
a>
E
a>
p
p
X
o
p
o
o
ttf)
X
i-H
P
43
bO
•rH
i-H
in
m
•rH
P
o
p
tn
S
in
•rH
P
43
P
X
o
p
o
E
o
P
o
•rH
P
•H
P
.
r--
oo
♦rH
i-H
od
X
i-H
o
E
o
p
p
X
o
o
p
•rH
P
cr
r\
g
•rH
P
o
p
a>
43
p
E
o
p
P
P
a>
••
P
o
43
P
•rH
a>
P
p
r\
E
o
"■p
o
o
_Q
H—*
r\
X
p
O
P
O
P
O
E
X
r-H
o
E
o
P
p
X
o
P
cti
43>
tn
•rH
s
o
o
p
m
o
i
p
r-H
<D
to
p
UJ
r\
o
p
•rH
r-H
&
•rH
o
m
•rH
P
1
P
r-H
<D
m
m
p
•H
+->
43
bO
■H
01
••
•*
+->
43
bO
••
rO
o
p
•H
P
cr
X
i"H
4)
E
o
P
p
X
o
p
o
O
bo
43
X
P
P
P
\
rX
P
o
O
+->
35
•H
cr
••
p
o
o
bo
bO
p
•rH
&
o
f-H
<D
>
o
Q
#
CO
o
o
bO
65
C
D
is:
of that material
CD
Ph
tn
X
P
p
H
nd
'S
X
X
CD
X
7— <
a)
CD
E
E
p
f-t
P
a)
CD
X
a)
and explanations
x
CD
CD
•rH
P
cr
a"
•
> P
rt o
^d X
P
X
O
p
o
p
p
a)
f-t
P
o
X c/3
p
CD O
rX -H
•H G
t
-H
* iH
Ph
TJ O
i-H
x
i— H
p
p
^d
x
on the tests
bO
bO
p
a)
o
tn
2
P
XZ
C
DP
i-H
PhP
O P
C
D cd
P
f-t
P
x
a)
X
p
p
on what will be covered
information
bJO
p
CD
•H
•H
P
id
CD
Getting
p
o
p
10.
X
o
p
X
r— I
P
CD
X
CD
i— H
p
x:
.beO
x
bO
tn
T3
P
ccj
C/3
p p
f-l
f-t O
<0 P
5-3
O
£
rX O
G i^f
•H
rP -tJ
P P
P
P
o tn
X P
P
I-H
o o
sd
CD
P
•H
P
tn
tn *h
P
CD
p
p
nd
P
•rH
P
cr
cr
x
r-H
a)
E
a)
f-t
p
x
i—i
a)
e
CD
P
P
X
x
a)
a)
TJ
o
o
bO
p
•H
P
O
ccj
o S
P o
O P
O
f-t
CD
X
P
O
,p
o
p sd
p p
C
D
T3
o
o
«P
tn p
P P
O <D
•H tn
P
tn
p
C
D
p b£)
cr p
•rH
<P C
D
o Xi
bO
bO
P
•rH
P
P
T3
X
P
Td
X
p
CD
>
CD
p
p
X -H
C
D T3
P P
C
D
C
DP
X P
H P
P
T3
tn
X
i-H
CD
E
CD
P
P
X
CD
••
CD
P
•rH
P
cr
tn
tn
P
r—1
o
tn
••
•rH
sd
X
i-H
p
sd
bC
•rH
i-H
p
nd
G
CD
+->
P
P
Td
I—1
p
o
sd
c/3
t-H
PC
P
tn
,,
P
CD
sd
p
•rH
CD
P
••
X
i-H
+->
sd
tuO
•rH
i-H
C/3
•rH
X
p
CD
E
o
•4
p
p
p
p
p
p
o
E
P
P
cr
•rH
••
•rH
CD
p
p
o
X
£
CD
i-H
Ph
O •
CD p
Ph o
P tn
t/3 in
O CD
s tn
X
r— 1
CD
E
P
P
P
X
P
X
i-H
<D
rH
p
•§
p
p
E
o
c/3
tn
p
X
i-H
p
M
•rH
i-H
p
P
tn
P
p
g
g
X
f-l
p
E
P
P
P
X
tn
tn
tn
P
i-H
o
C/3
•rH
P
••
x
P
P
•rH
P
cr
p
E
o
P
tp
p
X
i-H
p
rP
bO
•rH
i—t
G
P
tn
Xi
P
p
,P
nd
i-H
P
o
••
X
I--I
p
p
p
E
o
X
bO
•rH
i—t
tn
••
p
p
p
p
p
o
&
E
•rH
P
P E
c/3 g
o p
S W
X
fH
P
E
P
P
p
X
p
tn
p
X
p
bO
P
•rH
P
P
nd
X
cd
Td
X
p
p
>
p
••
p
p
•H
P
cr
••
X
i—t
p
tn
tn
x
bO
•H
i—t
P
T—H
o
tn
P
cr
••
X
r— 1
P
E
P
P
P
X
P
X
r— t
p
rX
•H
i-H
•rH
Xi
P
P
sd
p
•rH
P
P
p
nd
P
P
P
P
P
Td
rH
P
o
,G
.•
X
i-H
+->
X
bO
•rH
i-H
tn
t-H
tn
P
p
•rH
P
P
P
o •
xi P
5 o
•rH
p tn
i
h tn
Ot P
O C/3
P
Ot P
X
r—1
P
-X
•rH
i-H
P
P
tn
P
p
rP
P
•rH
P
P
tn
t-H
rX
P
•rH
xz
tn
..
rP
p
P
•H
f-t
P
X
i-H
CD
PC
•rH
i-H
c
tn
•*
.X
P
•rH
sd
P
p
•rH
P
cr
p
p
o
tn
tn
p
tp
o
P
Ph
P
O
P
O
P
P
P
tn
P
•H
X
s
to
**
X
i-H
P
E
P
P
p
X
•H
tn
tn
p
tn
P
p
g
§
P
tn
P
•rH
P
cr
P
sd
p
bO
p
•rH
p
p
Td
X
P
"G
X
P
P
>
•*
X
i-H
p
X
bO
T—♦
(/)
•H
P
P
P
X
P
•rH
P
P
tn
tn
P
i—t
O
••
rX
-H
+->
X
bO
•rH
i-H
tn
•rH
X
p
T3
V-t
P
P
P
cd
’id
i-H
P
o
sd
P
••
tn
••
P
p
•rH
P
cr
tn
X
rH
P
E
P
P
P
X
P
X
i-H
P
rX
•rH
i-H
•
P
o
X
i-H
p
,X
•H
i-H
e
p
i
—t
,X
P
••
•rH
X
p
r—s
tn
\___/
P
P
P
P
cd
Oh
X
s
X
i—t
p
E
P
P
P
X
P
X
i-H
P
rX
•rH
i-H
p
CD
CD
M
Pi
•rH
•rH
rH
CO
§
X
(“ 1
CD
s
CD
during
P
P
X
CD
day
CD
P
•rH
<u
Pi
•H
i-H
rH
c
3
CD
42
P
CD
43
P
X
CD
GO
S3
•H
P
e
CD
GO
c
•H
B
P
X
P
X
CD
P
X
CD
CD
P
3
X
CtJ
X
P
3
X
X
•H
3
cr
CD
>
CD
every
X
r-H
+
->
class
43
GO
•H
i-H
CO
CD
P
•H
X
3
cr
p
CD
>
this
P
CD
42
P
attend
•rH
I should
CD
S3
4->
43
GO
H
—I
o
Pi
GO
§
CO
•H
43
P
CO
X
S3
CD
P
P
ccS
>.
■§>
3
CO
•H
42
*
P
CCS
X
r-H
i-H
p
3
O
43
43
GO
•H
HH
•rH
3
cr
Pi
S3
•H
43
P
CO
•rH
3
cr
i-H
•rH
CD
£
CD
P
P
X
CD
P
P
X
CD
CD
B
*'
CD
P
•H
CD '
43
3
cr
P
GO
M
*
X
P
>
CO
CO
CCS
-H
CD
£
CD
P
P
•rH
P
ccj
i-H
CD
P
N
X
i—H
•rH
X
p
CO
CD
P
CCS
CD
P
X
•rH
3
cr
b
P
s
vO
x
s
CD
CD
P
CD
P
•H
3
cr
P
CD
•H
3
cr
3
cr
to
o
42
•rH
GO
•H
i-H
i-H
p
p
CD
43
P
•H
CD
CD
CO
43
CCj
P
CD
43
P
p
p
p
43
GO
CO
43
GO
43
GO
CO
<p
G
•H
CD
S3
CD
43
P
P
CD
42
P
C/>
•H
43
P
CD
S3
CD
43
P
•H
CD
S3
P
42
GO
•H
P
43
GO
•H
i-H
i-H
CO
CO
CD
P
CD
P
•H
•rH
g
P
3
O
43
co
P
43
GO
•H
P
S3
CD
CO
•8
i-H
co
CD
P
•H
i-H
3
3
cr
O
43
CO
p
CtJ
CD
p
p
Pi
p
X
CD
G
•H
X
CD
p
CD
CD
CD
X
CD
p
s
CD
o
CO
m
x
CD
B
p
•rH
P
X
CO
CD
Pi
CD
P
•H
S3
•rH
43
P
3
cr
P
43
•rH
3
cr
3
cr
Pi
CD
CD
S3
•H
43
P
b
b
3
O
43
CO
CO
CO
42
p
43
CD
Pi
•rH
P
CD
P
O
CD
fi
P
CD
P
CP
P
X
CD
X
p
o
p
P
CO
•rH
P
S
CD
P
•rH
to
co
CCJ
CD
£
CD
X
O
o
X
G
O
CO
e
p
p
CO
i-H
G
•H
CD
CD
Pi
B
CD
CD
e
CD
P
P
P
43
x
3
CD
P
P
CCS
i—l
CO
CD
>
B
*H
^
*
ccS
\
X
*H
P
CD
rS
^
G
o
§
i-H
CO
CD
43
GO
CD
CD
•rH
o
X
§
CD
Pi
i-H
.•
CO
CO
O
X
P
X
GO
•H
CO
43
CD
4->
S3
CD
P
X
CD
i-H
•rH
i-H
42
i-H
G
X
3
CD
P
P
CD
42
P
•rH
CD
P
P
B
o
CD
43
GO
•H
CD
G
*H
P
3
X
CD
Pi
•H
rH
•H
.iH
G
o
X
■
—
•
P
G
X
i-H
CD
G
3
i-H
43
CO
CO
thinks
husband/wife/fiance
My
LO
CD
43
P
ccS
X
i-H
x
CD
P!
•rH
rH
CO
CO
CD
CO
•i-l
CO
CO
P
CD
Pi
•H
§
to
CO
X
CD
CO
to
ccj
»—H
G
O
*H
S
P
p
3
CD
ccj
CD
X
cr
i-H
i—H
X
x
rH
the
summer
session.
66
o
3
p
p
CD
Pi
CD
CO
•rH
CO
S3
i-H
X
e
CD
Pi
X
s
CO
CD
CD
P
P
P
P
X
X
CD
CO
\— /
P
S3
CD
P
ccj
SP
X
s
CD
X
i-H
CD
CD
Pi
Pi
•rH
•rH
67
C-*
O
d
d
CD
iH
2
the
during
X
X
bO
d
•H
r-H
CD
CD
this
E
CD
X
0
d
d
r—
H
X
in
CD
CD
+->
•rH
d
cr
o
from
bO
•H
C/)
CD
+->
•rH
CD
be
|
bO
•rH
£
rH
V)
+->
••
0
fH
d
I should
X
rH
+->
X
+->
C/)
fH
CD
X
d
X
0
bO
•H
rH
c/)
••
CD
+->
•H
X
r-H
CD
X
•H
rH
d
•rH
C/)
X
tf)
r—
H
CD
+->
X
•H
+->
CD
bO
*H
rH
m
§
C/)
C/)
fH
0
d
X
0
CD
+->
•H
X
r—
H
r—
H
rH
CD
6
CD
o
C/)
C/)
C/)
fH
fH
fH
CD
CD
CD
p.
\
+->
X
CD
m
d
X
rH
e
X
bO
•H
rH
m
•H O
fH *H
«+H I/)
C/)
X CD
w
CD
X
•H
rH
0
X
0
X
•H
2
X
cr
3
O
X
X
r—
H
0
E
0
C/)
i— i
fH
X
X
d
•rH
X
0
X
X
r—
H
<4H
r—
H
<D
C/)
X
E
h
—1
X
bO
•H
X
+->
o
rH
m
0
0
X
d
cr
\
fH
o
•H
X
o
d
rH
d
X
fH
0
X
X
-rH
0
d
X
X
r—
H
C
/
)
d
o
X
X
d
•H
X
X
0
E
0
fH
0
X
d
cr
•H
tn
_
_t
'
X
X
d
X
P
bO
•H
d
o
X
r—
H
rH
X
d
X
fH
0
X
o
X
X!
d
X
o
X
X
rH
d
X
CD
X
•rH
r—
H
X
X
G
d
rH
X
0
X
o
X
•rH
0
d
O
X
X
X
bO
•H
d
o
X
rH
m
••
0
d
o
d
E
s
o
X
X
•H
d
cr
X
X
bo
d
•H
X
d
&
0
P
C/)
E
0
X
e^*
X O
<
—<d
rH
d d
rH
0 d
fH
d o
0 X
CJ t
n
LO
C
/
)
0
X
•rH
d
cr
S
o
X
rH
fH
rH
o
d
X
o
d
E
r*
bO
d
•H
X
d
0
0
bO
•H
d
o
X
X
0
X
o
d
E
s
o
X
E
0
fH
X
0
X
rH
rH
X
X
bO
•H
rH
C
/
)
••
0
X
•H
d
cr
••
X
r—
H
X
0
d
d
E
0
fH
X
X
0
0
P
C
/
)
X
rH
rH
d
fH
0
d
X
0
d
2
CJ
0
CJ
vO
X
i-H
0
0
d
•H
X
d
X
o
r—
H
bO
rH
X
d
X
r-H
C/)
d
•
X
G
d
2
s
d
o
s
.
O
O
0
X
d
cr
•H
•H
••
4h
X
•H
d
d
X
C
/
)
d
X
i
H
X
X
bO
•H
rH
in
••
X
0
X
fH
£
•rH
0
d
0
X
X
X
0
IX
\
'd
in
0
E
C
/
)
bO
••
X
r—
H
X
d
•H
X
X
0
o
d
d
£
•rH
rH
>
o
X
+
>
X
X
o
d
c
X
X
0
d
0
••
fH
X
X
d
o
X
X
X
0
••
,
—N
d
X
o
"d
T3
r—
H
d
o
X
>
£
0
X
C/)
G
•H
X
o
d
••
d
r-H
d
•
CD d
rH
X
CD
X
fH
X
X
0
O
XJ
T3
rH
d
X
o
d
E
X
fH
0
•
+->
r-H
CD
6
CD
0
fH
X
V)
X
G
••
cr
5
X
m
C/)
CD
d
d
rH
d
fH
C/)
•rH
o
f_,
m
+->
d
CD
d
X
cj
0
X
•H
X
••
E
fH
0
d
bO
•H
X
d
•rH
X
+->
p
E
•rH
X
rH
0
o
4h
O
••
f
s
.
>
fH
0
X
X
o
0
1
■x
X
0
X
(/)
(/)
fH
X
X
X
O
X
X
l
—H
0
E
0
X
o
d
o
X
r—
H
+->
X
o
X!
S
CD
X
•H
2
cr
r-H
d
d
X
•rH
d
+->
X
X
CD
••
•rH
X
X
o
d
cr
x
X
X
X
d
>
o
X
r-H
+->
d
E
d
cr
CD
E
CD
+->
•rH
CD
1—H
X
r-H
+->
rG
r—
H
m
0
d
o
X
3
X
+->
bO
d
C/)
X
X
rH
d
(/)
d
Q
X
fi
X
d
d
••
•H
:3
m
+->
X
X
fH
CD
X
u
d
X
d
■H
X
rH
(D
t
n
••
•rH
X
C/)
CD
CD
g
H
3
+->
d
+->
C/)
•H
rH
absent
E
C/)
r—
H
think
X
CD
E
o
X
m
rH
CD
••
•H
m
•rH
+->
CD
CT
My family/relatives/children
summer session.
d
d
d
+->
fH
class
sometimes
e
d
o
rH
fH
r-H
CD
•
•rH
•H
d
d
X
12.
r—
H
X
X
X
X
0)
x
CD
r—1
0
d
o
X
O
X
d
rH
rH
d
X
d
X
d
o
X
o
0
d
X
d
X
••
£
X
rH
o
d
o
X
X
X
bO
•H
X
in
rH
d
d
£
d
o
X
o
d
X
o
d
0
X
•H
d
cr
X
0
r—
H
E
E
0
X
X
0
o
X
X
•N
r—
H
d
X
0
d
0
bO
rH,
i—1
<T,
X
d
X
X
2
X
•H
d
IH
oo
rH
o
2
19.
In
general,
how
much
do
you
want
to
do
what
your
family/relatives/children
think
you
should
do?
68
Appendix C
70
0
CM r—1
P-i 0
•H >
a 0
r—1
o
cm
ft
p
p too
0 •H
•H to
CM
•H
P
toO
•H
CO
ft.
rH
nt
CM
vD
cn
rH
o
oo
•
*
to
(M
II
II
0
CM
O
&
P
co
p
•
P
*\
0
i-H
•
VI
o LO
o
t
0 f
tO
•
to
•
t— \
/— \
CM 1
'P 1
rH
'_/
CM 1
nd 1
nt
'— '
CM
M
to
0
CM
X
X
X
3
4->
E-h
X
!3
S Ch
•
•
CO
• to
to
*
•
p
p
p
p
•V
ft
ft. 'vf
ft
00
CM to vO LO
00 0> to o
rH
oo r H
v£>
rt-v.
a
•
•
•
to
• CM
rH
•
•
to
•
II
ii
II
II
•
V
V
•s
oo
o
CM
rH
CM
CM
X
CM
X
CM
X
H
ft.
o
LO
oo
to
ft
»
r^-
CM
rH
rH
CM
X
•
tO
•
P
ft,
00
CM
•
V)
V)
•
P
•v
•
P
ft,
rH
CM
P"
■P"
1
1
II
II
•
Vi
•
p
•S
a
CM
•
1
II
II
II
II
\ /— , ,— \
/— ,
CM 1 CM I cm I CM I CM 1
nd 1 nd nd nd j nd
r^.
rH vO CM r H
'— ' '__ / '— > '— '
'— '
X
•
P<|
f— \
CM
LO
o
•
f—
\
CM 1
nd |
LO
'— '
CM
X
/— \
/— ,
r— \
r— \
CM 1
nd
O
oo
CM
nd
o
oo
CM 1
nd
o
oo
CM
nd
o
00
to
'
M j
to
to
to
,__ /
' __ /
,__ /
M j
+->l
+->
rH
\D
LO
to CM
to to
nt LO
to to
•nf \D
to LO
'—
nd
0
CO
.§
P LO
CO CM
nt oo
• •
p- cCM
P" LO
X
P
nd
nd
p
CO
a.
rH
II
cr>
\o
•
00 CM
ft
•
rH
rH
LO CM
vO
00
•
•
LO
vO
•
oo ln
rH
rH
P 2
M
cj
P
0
r-H
o
LO P"
\0 rH
O ret
&<
S to
P
p
E -i
a.
cm
CO i— i
•
'vf CO
•
p- C"
CM
to oo
p
00
LO
tO
CM
•
LO
rH
CM LO
oo to
to to
cr>
00
LO
tO
•
CM
o
tO
nt
•
rH
rH
o
rH
X
ft f
t
•
LO
00 LO
rH i— l
0
<D
4->
0
CQ
tO
4->
P
0
p
o
nd
to
•H
&
4->
o
2
&
p
o
P
4->
CO
o
u
p
00
o
p
Cl,
0
0
P,
•H
ftp
P
00
to
p
4->
0
a
to
CD
P
p
0
r-H
M
tM
ftp
p
•H
to
.
to
>
.M S
p
P 0
X 2 tin
0
co
•H
-M
•H
CJ
co
1
—1 1
—1
0
p
p
0
p
M
-H
P
P
2
to
4->
O
0
•r—i
X>
P
co
0
0
co
CX
2
P
0
4->
P
P
nd
P
P
CJ
N
M
P
0
nd
P
+->
CO
0
+->
p
P
nd
P
p
toO
P
0
nd
P
z>
P
0
P
E-*
P
0
ftX
P
0
,P
E-
O
M
X
rH
4->
P
0
P
P
P
CJ
X
r-H
P
o
-H
4->
P
p
nd
P
p
Cd
Cm
O
r
P
P
0
X
X
i— 1
-H
B
P
CL,
cm
O
0
4->
P
0
P
P
P
CJ
nd
0
•M
0
0
Ph
X
nd
0
ftX
UJ
p
to
p
p
o
2
P
M P
P
-H P
£
O
0 Q 1 nd 0 Q
0 0 2 co 0 2 co
P
too
P
tH
<
CJ
o
to
p
p
o
2
M
-H
nd
0
P
CJ
2
to
P
P
0
2
P
P
o
a
co
,
2
•
o
2
P
P
0 Q
2 co
71
0)
O
P
0)
Pi
CD
PH
Ph
•H
O
r-H
CD
>
CD
i—1
•
lO
P=H
O LO
o
CD
•
O
P
•
to w>
O *M
•rH (/)
Ph
•rH
P
W)
•rH
C/D
•
•
to
•
P
P
•\
•sdtO
*
1
II
+->
to
CD
H
CD
i-H
Ph
£ ctj
TO LO
C/D CM
cn
r-H
o
LO
•
C7)
r-H
X
II
”0
3 2
4->
C/D
P
*\
00
tM
*
1
r
CNJ
i—1
•
I
II
II
/— \
Ph 1
T3
O
oo
to
v__>
■P i
PH
O
VI
/— \
Ph 1
t 3
©
oo
to
'__'
p 1
,— ,
Ph
TJ|
o
oo
to
’
P i
00
\o
O
•
LO
r-H
o
LO O
lo o
Oi O
•
•
CM r-H
oo
r—1 O
f"- \ D
lO
•
•
r-H
r-H
cm cn
r-H
OS O
•
•
CM r-H
00
cn lo
CD tO
vO vO
f
•
r-H
CD
rH
.
£ to
TO CM
C/D r-H
r-H
t o cn
^}" r-H
OD H
H-> II
O
'—i 2
•rH
Cl,
0 0 LO
r-H r-H
p
•
m
TO3
CD
PS
u
O
3=
CD
rQ
f—\
"TO
CD
P
p
*H
4->
p
o
o
V._/
r-H
cd
0)
i—1
rQ
TO
E-
o
+->
CD
+-»
O
CD
P h^
CD
,-H
rO
TO
•H
fH
TO
>
CD
TO
X
u
CO
M
P
O P
TO
2
CD Q
• 2
w
o
2
H
CD
<
CL,
CD
P
TO
CD Q
2
w
TJ
CD
+-» P
O
CD CD Q
P h S c/ d
x
PJ
TO
72
Table C2
M o d e l FA-1
(Prior b e h a v i o r d a t a omitted A T T § SN to INT § TB„
all p a t h s estimated s a t u r a t e d model)
M e a s u r e m e n t Model
Measure
F a c tors § V a r i a b l e s
S t a ndardized
F a c t o r Loadings
(A )
Unique V a r iance
(6 or £) 1-A2 (stand.)
Attitude
XI
X2
.708
.754
.499
.431
S u b j e c t i v e N o rms
X3
X4
.850
.849
.278
.278
I n tention
Y1
Y2
.826
.798
.317
.363
Target Behavior
Y3
Y4
.654
.661
.572
.563
C a usal Mod e l Para m e t e r s
Standardized Parameters
Factor Correlations
^ATT-SN
Y 11
Y 21
Y 12
Y 22
3 21
Residual Variances
Intention
T a r g e t Behav i o r
Standard
Weights
Critical Ratio
(t-values)
.542
5.278
.613
4.886
.211
1.334
-.012
-.116
-.124
-1.120
.501
3.670
.632
.628
5.168
3.392
73
Table C3
M odel F A - 2
(Prior b e h a v i o r d a t a omitted AT § SN to INT § AT
to target behavior)
M e a s urement Model
Measure
Fact o r s
§ Variables
S tanda r d i z e d
F a c t o r Loadings
(A)
Uniq u e Va r i a n c e
(6 o r e ) 1-A2 (stand.)
Attitude
XI
X2
.711
.753
.494
.434
S u b j e c t i v e Norms
X3
X4
.859
.840
.261
.295
Inten t i o n
Y1
Y2
.824
.801
.322
.359
Target Behavior
Y3
Y4
.671
.644
.549
.585
C ausal Model Parameters
Standardized Parameters
Factor Correlations
^ATT-SN
Y 11
Y 21
Y 12
Y 22
e 21
R esi d u a l V a r i a n c e s
Intention
Target B e h a v i o r
S tandard
Weights
C riti c a l Ratio
(t-values)
.539
5.199
.620
4.948
.121
.942
-.023
-.234
D E L ETED
.516
3.735
.631
.643
5.151
3,431
74
T able C4
Mo d e l F A - 3
(Prior b e h a v i o r d a t a o mitted paths A T T § SN to INT not to TB)
M e a s urement Model
Measure
F act o r s § V a r i a b l e s
S tandardized
Factor Loadings
(A)
Unique V a r i a n c e
(6 or c) 1-A2 (stand.)
Attitude
XI
X2
.713
.749
.492
.439
Subjective Norms
X3
X4
.859
.840
.263
.294
I n tention
Y1
Y2
.819
.801
.329
.358
T a rget B e h a v i o r
Y3
Y4
.655
.660
.571
.565
C ausal Model Parameters
Standardized Parameters
Factor Correlations
^ATT-SN
Y 11
Y 21
Y 12
Y 22
3 21
Residual Variances
Intention
Target Beha v i o r
Standard
Weights
Critical Ra t i o
(t-values)
.543
5.214
.633
5.039
DELETED
-.031
-.310
DELETED
.603
5.516
.619
.636
5.152
3.384
75
T able C5
C r i t i c a l R a t i o s of P a r a m e t e r s - - M o d e l BS-1
A ll P a ths E s t i m a t e d - S a t u r a t e d Mo d e l
M e a s u r e m e n t Model
Measure
Factors
£ Variables
S tandardized
F a c t o r Loadings
(A)
U n i q u e V a r iance
(6 or £ ) 1-A.2 (stand.)
Attitude
XI
X2
.716
.746
.487
.444
S u b j e c t i v e Norms
X3
X4
.879
.821
.228
.325
Prior Behavior
X5
X6
X7
.682
.662
.674
.535
.561
.546
Intention
Y1
Y2
.777
.849
.396
.279
T a r g e t Behav i o r
Y3
Y4
.616
.702
.621
.507
C a u sal Model Parameters
Standardized Parameters
Factor Correlations
^A-S
^A-PB
^S-PB
Y 11
Y 21
y 12
y 22
y 13
Standard
Weights
Critical Ratio
(t-values)
.331
5.152
.223
4.225
.015
.323
.390
2.268
-.077
-.552
.096
.947
.115
1.205
.556
4.252
76
T a b l e C5
(Continued)
Causal Model Parameters
S t a n d a r d i z e d Para m e t e r s
Fact o r C o r r e l a t i o n s
Y 23
321
Residual Variances
I ntention
T a rget B e h a v i o r
Standard
Weights
Critical Ratio
(t-values)
.934
5.644
.059
.550
.509
.161
4.984
1.273
77
Table C6
M o d e l BS-2
(Path SN to TB deleted)
M e a s u r e m e n t Model
Measure
Factors § V a r i a b l e s
Standardized
F a c t or Loadings
(X)
Unique Va r i a n c e
fo or e ) 1-A2 (stand.)
Attitude
XI
X2
.712
.749
.493
.438
S u b j e c t i v e N orms
X3
X4
.862
.837
.257
.300
Prior Behavior
X5
X6
X7
.683
.664
.679
.534
.559
.539
I n tention
Y1
Y2
.778
.847
.394
.282
Target B e h a v i o r
Y3
Y4
.611
.708
.627
.499
Causal Model P a r a meters
Standardized Parameters
Factor Correlations
^A* S
^A-PB
^SN-PB
Yll
Y21
Y1 2
Y 22
^13
Standard
W eig h t s
Critical R a t i o
(t-values)
.544
5.243
.437
4.196
.050
.611
.355
2.824
.028
.251
.082
.851
DELETED
.433
4.304
78
T a b l e C6
(Continued)
C a u sal Model Parameters
Standardized Parameters
Fa c t o r C o r r e l a t i o n s
y 23
21
Residual Variances
Inte n t i o n
Ta rget B e h a v i o r
Standard
Weights
C r i tical Ratio
Cl-values)
.833
5.960
.086
.674
.510
.190
4.994
1.528
79
Table C7
Model BS-3
(Paths SN to TB, A T T to TB deleted)
M e a s u r e m e n t Model
Measure
Factors § Variables
St a n d a r d i z e d
F a ctor Loadings
(A)
Unique V a r i a n c e
(6 or e ) 1-A2 (stand.)
Attitude
XI
X2
.713
.749
.492
,438
Su b j e c t i v e Norms
X3
X4
.861
.838
.258
.298
Pr i o r B e h a v i o r
X5
X6
X7
.682
.664
.679
.535
.559
.539
I n tention
Y1
Y2
.779
.846
.393
.284
Target Behavior
Y3
Y4
.609
.710
.629
.497
Ca usal M o d e l Parameters
Standardized Parameters
Fa c t o r C o r r e l a t i o n s
^AT-SN
^AT-PB
^S N - P B
Yll
*21
*12
*22
*13
Stand a r d
Weights
Critical Rat i o
(t-values)
.543
5.245
.441
4.312
.054
.655
.355
2.812
DE L E T E D
.082
.857
DE L E T E D
.431
4.278
80
T a b l e C7
(Continued)
C a usal Mod e l Parameters
Standardized Parameters
F a ctor Corx'elations
Y 23
21
Residual Variances
Intention
T a rget B e h a v i o r
Standard
Weights
Critical Ratio
(t-values)
.835
6.003
.103
.943
.512
.190
5.017
1.527
81
Ta b l e C8
M o d e l BS-4
(All p a t h s to TB from e x o g enous v a r i a b l e s deleted)
M e a s u r e m e n t Model
Measure
Factors § Variables
Standardized
F a c t o r Loadings
Uni q u e V a r i a n c e
CO
fc or e ) !-X2
Attitude
XI
X2
.715
,747
.448
.443
S u b j e c t i v e Norms
X3
X4
.875
.825
.234
.320
Prior Behavior
X5
X6
X7
.690
.639
.687
.524
.591
.528
Intention
Y1
Y2
.745
.825
.445
.320
Target Behavior
Y3
Y4
.643
.673
.587
.548
(stand.)
C a u s a l Mo d e l P a r ameters
Standardized Parameters
Factor Correlations
^ AT •SN
4>AT*PB
SN *PB
Y 11
Y 21
Y 12
Y 22
Y13
St a n d a r d
W e i ghts
Critical Ratio
(t-values)
.542
5.174
.444
4.223
.029
.340
..304
2.463
DELETED
.110
1.164
DELETED
.556
5.245
82
Ta b l e C8
(Continued)
Causal Model Parameters
Standardized Parameters
Factor Correlations
y
23
B21
Re s i d u a l V a r i a n c e s
In tention
T a rget B e h a v i o r
Standard
Weights
Critical Ratio
(t-values)
DELE T E D
.714
6.783
.397
.490
4.432
3.155
83
Table C9
Mo d e l BS-5
(F-A Models o n l y p a t h s from A T T § SN to INT § from INT to
targ e t b e h a v i o r - - p r i o r b e h a v i o r not inclu d e d except
in m e a s u r e m e n t model)
M e a s u r e m e n t Model
Measure
Fa c t o r s § V a r i a b l e s
Standardized
F a c t o r Loadings
(X)
Unique Variance
(6 or e ) 1-A2 (stand.)
Attitude
XI
X2
.615
.637
.621
.594
S u b j e c t i v e N orms
X3
X4
.873
,822
.238
.324
Pr i o r B e h a v i o r
X5
X6
X7
.683
.644
.690
.533
.586
.524
Intention
Y1
Y2
.762
.824
.419
.321
Target Behavior
Y3
Y4
.646
.669
.582
.553
C a u sal Mod e l Parameters
Standardized Parameters
Factor Correlations
^AT-SN
^AT-PB
^SN-PB
■Yll
^21
^12
y 22
y 13
Stand a r d
Weights
Criti c a l Ratio
(t-values)
.646
5.481
.607
4.934
.022
.261
1.035
6.302
DE L E T E D
-.395
DE L E T E D
DE L ETED
-2.984
/
84
T able C9
(Continued)
Causal Mo d e l Parameters
Standardized Parameters
Factor Correlations
y 23
3 21
R esidual V a r i a n c e s
Inten t i o n
Target Behavior
Standard
W e i ghts
C ritical Rati<
(t-values)
D ELETED
.687
6.480
.309
.527
2.995
3.254
85
T able CIO
M odel BS-6
(Prior b e h a v i o r to INT o n ly A T T to INT £ TB,
SN to INT, PB to INT)
M e a s u r e m e n t Model
Measure
Fact o r s
§ Variables
Standardized
F a c t o r Loadings
a)
Unique Variance
(6 or e ) 1 - x 2 (stand.)
Attitude
XI
X2
.713
.747
.491
.442
S u b j e c t i v e Norms
X3
X4
.876
.823
.282
.322
Prior Behavior
X5
X6
X7
.689
.639
.688
.525
.592
.527
In tention
Y1
Y2
.749
.835
.439
.303
Targ e t B e h a v i o r
Y3
Y4
.650
.665
.577
.558
Causal Model Parameters
St a n d a r d i z e d P a r a m e t e r s
Fa c t o r C o r r e l a t i o n s
^ATT-SN
^ATT-PB
^SN-PB
*11
*21
*12
*22
*13
Standard
Weights
Critical Ratio
(t-values)
.540
5.159
.457
4.321
.029
.339
.283
2.239
.080
.654
.116
1.209
DELETED
.548
5.068
86
T a b l e CIO
(Continued)
C a u sal Mod e l Parameters
Standardized Parameters
Factor Correlations
y
23
3 21
Residual Variances
I ntention
Targ e t B e h a v i o r
Standard
Weights
C r i tical Ratio
(t-values)
DELETED
.649
4.777
.426
.510
4.5 5 4
3.238
87
Table Cll
Model BS-7
(No p a t h s f r o m p r i o r b e h a v i o r p a t h from A T T to TB^
B e n t l e r - S p e c k a r t M o d i f i c a t i o n 1)
M e a s u r e m e n t Model
Measure
Factors
§ Variables
S t a n dardized
F a c t o r Loadings
M
U n i q u e Variance
(6 or e ) 1-A2 (stand.)
Attitude
XI
X2
.610
.634
.628
.597
Subjective Norms
X3
X4
.892
.808
.205
.347
P r i o r B e ha v i o r
X5
X6
X7
.678
.649
.689
.540
.578
.525
Intention
Y1
Y2
.770
.853
.408
.273
Target Behavior
Y3
Y4
.680
.636
.537
.596
Causal M o d e l P a r a m e t e r s
Standardized Parameters
Factor Correlations
^ATT-SN
^ATT-PB
^S N - P B
Y 11
Y 21
Y 12
y 22
y 13
S tand a r d
Weights
Critical Ra t i o
(t-values)
.575
4.956
.695
5.416
.025
.296
.844
6.409
.500
3.145
-.217
-2.086
DELETED
D E L ETED
88
Ta b l e Cll
(Continued)
Causal Mo d e l Parameters
Standardized Parameters
Factor Correlations
y 23
321
Residual Variances
I n tention
Target Behavior
Standard
Weig h t s
'
Criti c a l Ratio
(t-values)
DE L ETED
*260
1.770
.451
.496
4.199
3. 249
Appendix
90
D i s c u s s i o n of Path An a l y s i s
Path analy s i s
is the logical p r e c u r s o r of structural analysis.
Th i s d i s c u s s i o n is b a s e d u p o n the fol l o w i n g references:
Wroten,
1975;
1978;
C o o k § Campbell,
and N a m b oodiri,
Carter,
1979;
Billings §
K e r linger § Pedhazur,
1973;
Li,
§ Blalock,
1975.
f
In o r der to use p a t h a nalysis the t h e oretical fra m e w o r k b e i n g
s t u d i e d m u s t first be m a d e explicit.
The basic t e c hnique uses ordin a r y
least squares r e g r e s s i o n to der i v e p a t h coeffi c i e n t s w h i c h m a y be defined
as s t a n d a r d i z e d r e g r e s s i o n c oe f f i c i e n t s and w h i c h m a y be i n t e rpreted as a
r a t i o o f two s t a n d a r d deviations.
These p a t h c o e f f i c i e n t s b e t w e e n v a r i ­
ables are t h e n u s e d to test p r o p o s e d causal r e l a t i o n s h i p s am o n g a set of
va r iables.
In this manner,
p a t h analysis can be u s e d to test an a p r i o r i
causal h y p o t h e s i s a g a i n s t a set of ob s e r v e d correlations.
Li
(1975)
techniques
the u s e f u l n e s s o f p a t h analysis over simple mu l t i p l e r e g r e s s i o n
lies in its a b i l i t y to ext e n d the single m u l t i p l e r e g r e s s i o n
e q u a t i o n t r e a t m e n t to a n e t w o r k of v a r i a b l e s
equation.
A c c o r d i n g to
The use o f p a t h analysis
inv o l v i n g m o r e than one
for d e c o m p o s i n g a c o r r e l a t i o n into its
c o m p o n e n t s of direct a n d indirect effects w i t h i n a given causal model
an a d d i t i o n a l
i mportant f u n c t i o n of this analytic approach.
Kerlinger and Pedhazur
(1973)
expl a i n the use of p a t h analysis as
a tool for t h e o r y t e s t i n g as follows:
P a t h a nalysis
testing.
is
is an i m portant analytic tool for t h e o r y
T h r o u g h its a p p l i c a t i o n one can d e t e r m i n e w h e t h e r
or not a p a t t e r n o f c o r r e l a t i o n s
for a set o f ob s e r v a t i o n s
is c o n s i s t e n t w i t h a s p e c i f i c theoretical
formulation
. . . .
a c o r r e l a t i o n b e t w e e n two v a r i a b l e s can'be exp r e s s e d
as a c o m p o s i t e of the direct and indirect effects of one
v a r i a b l e on the other.
U s i n g p a t h c o e f ficients it is t h e r e ­
fore p o s s i b l e to r e p r o duce the c o r r e l a t i o n m a t r i x
the v a r i a b l e s
in the system.
. . . however,
(R)
for all
as long as all
v a r i a b l e s are c o n n e c t e d b y paths and all the p a t h coeff i c i e n t s
are employed,
the R m a t r i x can b e r e p r o d u c e d r e g a rdless of
the causal model f o r m u l a t e d b y the researcher.
Consequently,
the r e p r o d u c t i o n of t h e R. m a t r i x w h e n all the p a t h co e f f i c i e n t s
are used
is of no h e l p in t e s t i n g a specific theoretical model.
W hat if one w e r e to d e le t e cert a i n paths from the causal m o d e l ?
This,
in effect,
will amount to sett i n g c e r t a i n p a t h c o e f f i ­
cients euqal to zero.
The i m p l i c a t i o n is that the r e s e a r c h e r
conce i v e s o f the c o r r e l a t i o n b e t w e e n the two var i a b l e s whose
c o n n e c t i n g p a t h is d e l e t e d as b e i n g due to indirect effects
only.
By d e l e t i n g c e r t a i n paths the r e s e a r c h e r is o f f e r i n g a
mo r e p a r s i m o n i o u s
p aths,
causal model.
If after the d e l e t i o n of some
it is p o s s i b l e to r e p r o d u c e the original R
c l o s e l y a p p r o x i m a t e it,
correlations
ous model.
matrix,
the c o n c l u s i o n is that the p a t t e r n of
in the d a t a is consistent w i t h the m o r e p a r s i m o n i ­
. . .
If a f t e r the d e l e t i o n o f some paths there are large d i s ­
c r e p a n c i e s b e t w e e n t h e origi n a l R m a t r i x and the r e p r o d u c e d
one,
the c o n c l u s i o n is that,
a m o n g the variables,,
tenable.
or
(p. 317)
in the light of the relations
the mo r e p a r s i m o n i o u s theo r y is not
92
P ath a n a l y s i s r e q u i r e s a d i s t i n c t i o n b e t w e e n v a r i a b l e s
e x o g e n o u s and e n d o g e n o u s .
labeled
E xo g e n o u s v a r i a b l e s are one or m o r e v a r i a b l e s
w h o s e causes lie o u t s i d e the causal
be c o r r e l a t e d w i t h n o causal
system.
Variables
d i r e c t i o n specified.
arc those w h ose cau3e3 lie w i t h i n the system*
so labeled m a y
E n d o g e n o u s v a r i ables
These v a r i ables musl be
s p e c i f i c a l l y o r d e r e d w i t h r espect to h y p o t h e s i z e d cause-effect r e l a t i o n ­
ships.
The v a r i a n c e of an e n d o g e n o u s v a r i a b l e
is c o n s i d e r e d to be
a c c o u n t e d for b y the e f f e c t s of other e n d o g e n o u s and e x o g enous v a r i a b l e s
p r i o r to it in the o r d e r i n g and b y a residual or error term.
For each e n d o g e n o u s v a r i a b l e it is n e c e s s a r y to create a causal
model w h i c h is a w e i g h t e d f u n c t i o n of var i a b l e s p r i o r to
that v a r i a b l e a n d an e r r o r term,
(causal to)
Actual weig h t s are d e t e r m i n e d b y m e a n s
of m u l t i p l e r e g r e s s i o n in w h i c h each e n d o genous v a r i a b l e is trea t e d as
the c r i t e r i o n and the v a r i a b l e s h y p o t h e s i z e d to affect it are treate d
as p r e d i c t o r s .
The r e s u l t i n g s t a n d a r d i z e d b e t a w e i g h t s are p a t h c o e f ­
f i c i e n t s r e p r e s e n t i n g the d i rect effects of the causal v a r i a b l e s u p o n the
c r i t e r i o n v a riable.
A n indirect p a t h is computed by m u l t i p l y i n g toget h e r
all the direct p a t h c o e f f i c i e n t s w h i c h compr i s e the indirect path.
A b a s i c t h e o r e m of p a t h analy s i s is that "the c o r r e l a t i o n b e t w e e n
two v a r i a b l e s
able s "
(Lij
is the sum o f all c o n n e c t i n g pa t h s b e t w e e n the two v a r i ­
1975^ p.
b e t w e e n total e f f e c t s
ficients) .
149).
Therefore^
(the correlation)
indirect e f f e c t s are the diffe r e n c e s
and direct e f f e c t s
(path c o e f ­
In this way^ p a t h a n a lysis m a y be u s e d to d e c o m p o s e a
c o r r e l a t i o n into its c o m p o n e n t s of direct and indirect effects w i t h i n
a causal m o d e l .
93
In t e s t i n g a t h e o r y w h i c h p r e d i c t s the absence of one or more
di r e c t paths,
t hose v a r i a b l e s h y p o t h e s i z e d to have o n l y indirect effects
are d e l e t e d f r o m the r e g r e s s i o n e q u a t i o n for that p a r t i c u l a r dependent
v aria b l e .
If the h y p o t h e s i z e d p a t h s are s t a t i s t i c a l l y or m e a n i n g f u l l y
significant
(beta w e i g h t s u s u a l l y g r e a t e r than
.05)
able to r e p r o d u c e the o b s e r v e d c o r r e l a t i o n matrix,
and the model
is
the t h e o r y as m o d e l e d
is supported.
The m a j o r a s s u m p t i o n s that m u s t be met in the u s e of p a t h analysis
are:
(a) a p r i o r i s p e c i f i c a t i o n of the causal
(b) the s p e c i f i e d causal
(c)
the r e s i d u a l
sequence of the variables;
s e q uence m u s t be a one- w a y flow
(recursive);
of each e n d o g e n o u s v a r i a b l e affects o n l y that s p e cific
v a r i a b l e and is u n c o r r e l a t e d w i t h ot h e r system var i a b l e s or w i t h their
r e s iduals;
and
(d) the d a t a are
and on a r a t i o or interval
T here are,
analysis.
however,
linear,
additive
(no i n t e r a c t i o n effects),
scale of measurement.
some b a s i c s h o r tcomings
in the use of p a t h
The d e g r e e o f r e l i a b i l i t y w i t h w h i c h latent v a r i a b l e s or
c o n s t r u c t s are m e a s u r e d b y the o b s e r v e d v a r i a b l e s
is not a s s e s s e d wh e n
m u l t i p l e r e g r e s s i o n e q u a t i o n s u s i n g one ob s e r v e d i n d i c a t o r for each latent
v a r i a b l e are u s e d to o b t a i n the b e t a weights
o f this,
any e r r o r in m e a s u r e m e n t
(path c o e f f i c i e n t s ) .
Because
is c o n f o u n d e d with e s t i m a t i o n o f causal
p a r a m e t e r s p e r t a i n i n g to the latent variables.
Also,
the a s s u m p t i o n o f a r e c u r s i v e or u n i d i r e c t i o n a l
causal m o d e l w h i c h m a y
not always r e p r e s e n t reality.
the m e t h o d requires