Georgia Southern University
Digital Commons@Georgia Southern
SoTL Commons Conference
SoTL Commons Conference
Mar 12th, 4:00 PM - 5:45 PM
Difficulty Components in French Verb Tenses
Imparfait and Passé Composé for Anglophone
Learners
Ofelia Nikolova
Valdosta State University, [email protected]
Follow this and additional works at: http://digitalcommons.georgiasouthern.edu/sotlcommons
Part of the Curriculum and Instruction Commons, Educational Assessment, Evaluation, and
Research Commons, Educational Methods Commons, Higher Education Commons, and the Social
and Philosophical Foundations of Education Commons
Recommended Citation
Nikolova, Ofelia, "Difficulty Components in French Verb Tenses Imparfait and Passé Composé for Anglophone Learners" (2009).
SoTL Commons Conference. 124.
http://digitalcommons.georgiasouthern.edu/sotlcommons/SoTL/2009/124
This presentation (open access) is brought to you for free and open access by the Programs and Conferences at Digital Commons@Georgia Southern.
It has been accepted for inclusion in SoTL Commons Conference by an authorized administrator of Digital Commons@Georgia Southern. For more
information, please contact [email protected].
Difficulty Components of a French Grammar Test
Ofélia R. Nikolova
Valdosta State University
Abstract
Aspect Hypothesis
Method cont’d
This study examined tests of two French verb tenses –
imparfait and passé composé - the usage of which
represents a problem area for most native speakers of
English. Twelve cognitive operations were hypothesized,
which were proven to explain the difficulty of the test items
that were calibrated with the linear logistic test model. The
results of the study brought evidence that cognitive
categories, such as "action in progress interrupted by a onetime event" and "achievement" were significant predictors of
item difficulty and item easiness, respectively.
The information about the psychometric structure of a
grammar test and the particular "weights" of its components
in a model predicting item difficulty will allow test developers
to construct items whose difficulty is known prior to
administering the test without having to pilot individual items
in study groups. Knowledge about the cognitive operations
underlying item difficulty can help teachers target specific
cognitive and processing characteristics of the student
response style in a variety of subject matters.
"First and second language learners will initially be
influenced by the inherent semantic aspect of verbs or
predicates in the acquisition of tense and aspect markers
associated with or/affixed to these verbs" (Andersen & Shirai,
1994, p. 133).
Instrument and procedure
The subjects were given two cloze tasks – two texts (Appendix
1 & 2) in which the verbs were supplied in parentheses in the infinitive
and blanks were left immediately preceding the infinitive. The
participants were instructed to fill in the correct verb form – passé
composé or imparfait for each verb. Three native speakers of French
were given the same task to establish a baseline of native-like
answers.
Based on Vendler's four categories and learner think-aloud
protocols, the following cognitive operations were identified which led
the learners to chose one tense over another:
Background
The correct usage of the two past tenses passé
composé and imparfait is one of the highlights of mastering
the French language, allowing the learner to narrate and
describe in the past and it is at the same time a stumbling
point for native speakers of English attempting to learn
French.
Sources of Difficulties
•Aspectual rather than tense difference
•Lack of perfect match in translation
•Learner preference for one grammar aspect over another
based on the lexical aspect of the verb
•Complex, long, and confusing explanations in traditional
textbooks
Lexical Aspect - the Four Vendler Verb Categories
Grammatical aspect is external to the verb, lexical
aspect has to do with the meaning of the verb. Vendler
(1967) came up with four distinct categories corresponding to
the four possible lexical aspects for verbs:
•States: Je suis heureuse. (I am happy)
•Activities: Je mange. (I eat)
•Accomplishments: Je mange une pomme (I eat an apple)
•Achievements: J’entre dans la salle. (I enter the room).
Correct, native-like usage of grammatical aspect
evolves from prototypical to non-prototypical combinations in
eight stages (Andersen 1991).
1)No marks for tense nor aspect.
2)Perfective with punctual verbs only.
3)Imperfective with stative verbs only.
4)Add accomplishments for perfective and activities for
imperfectives.
5)Accomplishments with both markers.
6)Activities with both markers.
7)Achievements with both markers.
8)States with both markers.
However, verbal morphology is the end-point of the
development of expression of temporality. At the beginning,
learners resort to other devices, namely pragmatic and
lexical ones such as scaffolding, implicit reference,
contrasting events, and chronological order (Schumann,
1987), use of base or default form of the verb (infinitive for
French), connectives such as and, because, and so and
adverbials.
Literature Review
Overall, studies of passé composé / imparfait
acquisition have shown that it is a grammatical topic, which is
difficult to master for anglophones and that learners are
influenced by the lexical aspect of the verb in their choice of
grammatical form. No studies have conducted, however, a
testing of item difficulty of a passé composé / imparfait
grammar task, based on the psychometric structure of the
items, with a relatively large number of participants. The
present investigation is devised with the intention to bridge
this gap.
Method
Participants
The participants were 205 second-semester students of
French from several mid-western universities. However, the
number of participants was brought down to 146 because of
the missing data in many of the participant responses.
Table 1
Cognitive Operations Hypothesized to Explain Item Difficulty for the French Grammar Test
_____________________________________
1.States
2.Activities
2.Accomplishments
4. Achievements
5. Imparfait
6.Passé composé
7.Precise moment of event
8. Precise span of time
9.Frequency of repetition in the past
10.Action in progress interrupted by one-time event
11.Event interrupting action in progress
12. Agreement of tenses
There were a total of 62 tokens of verb occurrences in the two
texts. In order to avoid redundancy 25 verbs were selected to
participate in the study out of the total of 62. For each verb the
cognitive operations necessary to arriving at the native-like solution
were marked with the number one and the cognitive operations which
did not participate with the number zero. The correct (native-like)
answer for each participant was marked with a one and the incorrect
answer with a zero. If the subject did not give an answer, the mark
was nine.
The Linear Logistic Test Model (LLTM)
Results
In this study, the weight matrix of 25 items and their weights on
12 cognitive operations is a two-way (25 x 12) table with an entry of 1
when the item requires the respective cognitive operation, 0 otherwise
(Table 2). The LLTM basic parameters, j, in Equation 2 are
analogous to regression coefficients when item difficulties are
predicted from variables specified in the weight matrix. These
parameters are estimated here using the computer program LPCMWIN (Fischer & Ponocny-Seliger, 1998). This program provides also
statistical and graphical tests for the validity of the weight matrix with
the LLTM.
Table 3
Basic Parameters for Twelve Cognitive Operations Hypothesized to Explain the Difficulty of Items
in the French Grammar Test
_____________________________________________________________
Basic parameters
Standard error z-value
_______________________________________________________________
1
-0.4937
0.0841
5.87**
2
-1.3157
0.1658
7.94**
3
0.3319
0.1070
3.10**
4
1.0913
0.1122
9.73**
5
0.4709
0.0858
5.49**
6
-0.4709
0.0858
5.49**
7
0.4042
0.1157
3.49**
8
-0.3862
0.1605
2.41*
9
-0.5930
0.1979
2.99**
10
-1.5891
0.1450
10.91**
11
-.07406
0.1486
4.98**
12
-0.8583
0.1479
5.80**
______________________________________________________________
*
TEMPLATE DESIGN © 2008
www.PosterPresentations.com
p < .05. **p < .01.
Results (cont’d)
In Table 3, positive estimates of basic parameters
indicate that the cognitive operations associated with these
parameters increase the item difficulty. Conversely, negative
parameter estimates indicate that the cognitive operations
associated with these parameters facilitate the correct
answer on the respective items. Both Rasch and LLTM
estimates of the item difficulties for the test used in this study
are provided in Table 4 (higher algebraic values indicate
higher item difficulty).
Table 4
Rasch Item Difficulties and Their Predicted (LLTM) Values
______________________________________________________
Item LLTM
Rasch Item
LLTM
Rasch
____________________________________________________
1
0.0207 0.3972 14.
1.0679
0.6190
2
0.0207 -0.3671 15
-0.5212 -0.0486
3
1.0679 0.6672 16
-0.2161 -0.2436
4
-0.7429 -1.5833 17
0.3273 0.2377
5
-0.0768 -0.0486 18
-0.7429 -0.4274
6
-0.0966 0.7169 19
-0.9212 -1.4320
7
-0.0121 -0.6325 20
-0.9212 -1.4918
8
-0.8377 -0.3059 21
1.0679 0.9939
9
0.6638 0.8766 22
-0.6886 -0.7184
10
0.6638 0.8766 23
-0.8014 -0.8318
11
0.8638 0.5722 24
0.4248 0.4393
12
0.6638 -0.1151 25
0.0207 0.6190
13
-0.0956 0.3972
_______________________________
Note. The Pearson correlation between actual (Rasch) and predicted (LLTM)
item difficulties is r = 0.794.
Conclusions
All 12 difficulty components used by LLTM in this study
were statistically significant which validates their role in the
hypothesized cognitive structure of French grammar tests.
The complexity/ easiness components were strong predictors
of item difficulty or easiness because they were used in their
most non-prototypical or prototypical contexts respectively.
A new element in this study was the hypothesizing of
several adverbials as components of complexity for the
reviewed items. The adverbials were (7) "precise moment of
event indicated", (8) "precise span of time indicated", and (9)
"frequency of repetition in the past indicated". According to
the calculated difficulty of test items, out of 12 cognitive
operations, (7) came third easiest, (8) – fifth easiest, and (9) fifth most difficult. Why were in our case some adverbials
(7,8) more helpful than others (9) ? We believe that
adverbials pointing at the choice of passé composé have
been of better use for the learner because of the earlier
acquisition of passé composé. On the other hand, adverbials
referring to imparfait were less helpful because of the greater
internal complexity of the tense (Andrews, 1992; Kaplan,
1987).
Contact information
Ofélia R. Nikolova,
Associate Professor of French and Director,
University Honors Program
Valdosta State University
1500 N. Patterson
Valdosta, GA 31605
Ph. 229-249-4894
[email protected]