ABSTRACT (1 st )

WASON CARD SORT:
DATA ANALYSIS
Week 3 Practical
WASON CARD SORT
LECTURE 1
WEEK 3 PRACTICAL
PRACTICAL
WEEK 1
NO LECTURE
WEEK 2
NONPARAMETRICS 1
1ST PRACTICAL
WEEK 3
NONPARAMETRICS 2
1ST ANALYSIS
WEEK 4
SAMPLING DISTRIBUTIONS
1ST ANALYSIS + PROBLEMS 1
WEEK 5
HYPOTHESIS TESTING
2ND PRACTICAL
WEEK 6
RELATED T-TEST
2ND ANALYSIS + SOLUTIONS 1
WEEK 7
INDEPENDENT T-TEST
2ND ANALYSIS + PROBLEMS 2
WEEK 8
INDEPENDENT ANOVA
3RD PRACTICAL
WEEK 9
DEPENDENT ANOVA
3RD ANALYSIS + SOLUTIONS 2
NO LECTURE
3RD ANALYSIS + PROBLEMS 3
WEEK 10
NO PRACTICAL
WASON CARD SORT
LEARNING OUTCOMES
BY THE END OF THE SESSION, YOU SHOULD BE ABLE TO:
Use SPSS to test the first experimental hypothesis of the Wason
card sorting experiment and produce a related graph.
Use SPSS to test the second experimental hypothesis of the
Wason card sorting experiment and produce a related graph.
Use the graphs to interpret your chi-square findings.
Make a start on writing up your RESULTS and DISCUSSION
sections for your lab report.
WASON CARD SORT
METHOD RECAP
DESIGN
Half of the people in the room did the concrete first, half did the abstract.
ABSTRACT then CONCRETE
CONCRETE then ABSTRACT
HYPOTHESES
Our current experiment raises two hypotheses:
Q1: Is performance better on some versions of the Wason card
sorting task than others the first time it is performed?
ABSTRACT then CONCRETE
CONCRETE then ABSTRACT
Q2: Is performance on the abstract task affected if it follows a
concrete scenario?
ABSTRACT then CONCRETE
CONCRETE then ABSTRACT
WASON CARD SORT
The first column shows subject
number. And yes, there are 180.
RESULTS
The third column
shows whether the
individual got their
first problem right
or wrong.
1 = RIGHT
2 = WRONG
The second column
shows which type
of problem each
individual solved
first.
1=
ABSTRACT
2=
CONCRETE
The fourth column
shows whether the
individual got the
abstract problem
right or wrong.
1 = RIGHT
2 = WRONG
WASON CARD SORT
RESULTS
Let’s make the data look a little more meaningful by changing the numeric
values into textual values. Go to variable view and select values.
For first_problem, correct1 and correcta, associate the values (e.g., 1, 2) with
the appropriate value labels (e.g., abstract, concrete; wrong, right).
first_problem
1 = ABSTRACT
2 = CONCRETE
correct1
1 = RIGHT
2 = WRONG
correcta
1 = RIGHT
2 = WRONG
WASON CARD SORT
RESULTS
Q1: Is performance better on some versions of the Wason card
sorting task than others the first time it is performed?
CONCRETE (1st) vs. ABSTRACT (1st)
If we are interested in comparing
categorises of responses, then the
chi-square test is appropriate.
In SPSS, the chi-square test is
hidden away underneath
descriptive statistics > crosstabs.
Let’s go there now.
WASON CARD SORT
RESULTS
Q1: Is performance better on some versions of the Wason card
sorting task than others the first time it is performed?
CONCRETE (1st) vs. ABSTRACT (1st)
In order to build a chi-square table,
we need to put our various
categories into rows and columns.
Let’s put first_pr as a row and correct
as a column. This will show us the
frequency distributions.
Under statistics, we also need to
make sure the chi-square test is
performed, so tick that.
Under cells, also make sure that both
observed and expected are clicked.
WASON CARD SORT
RESULTS
Q1: Is performance better on some versions of the Wason card
sorting task than others the first time it is performed?
CONCRETE (1st) vs. ABSTRACT (1st)
CORRECT1 * FIRST_PR Crosstabulation
FIRST_PR
1
CORRECT1
right
wrong
Total
Count
Expected Count
Count
Expected Count
Count
Expected Count
2
8
20.5
82
69.5
90
90.0
33
20.5
57
69.5
90
90.0
Total
41
41.0
139
139.0
180
180.0
The second table (after case
processing summary) confirms
our 180 observations and displays
the frequency distribution of right
and wrong responses for the two
kinds of Wason card sort test.
Chi-Square Tests
Pears on Chi-Square
Continuity Correctiona
Likelihood Ratio
Fisher's Exact Test
Linear-by-Linear
Ass ociation
N of Valid Cas es
Value
19.740b
18.193
20.887
19.631
df
1
1
1
1
Asymp. Sig.
(2-s ided)
.000
.000
.000
Exact Sig.
(2-s ided)
Exact Sig.
(1-s ided)
.000
.000
The third table provides us with
the chi-square value, which may
be reported as:
.000
180
a. Computed only for a 2x2 table
b. 0 cells (.0%) have expected count les s than 5. The minimum expected count is
20.50.
χ2 (1) = 19.74, p < .001
WASON CARD SORT
RESULTS
Q1: Is performance better on some versions of the Wason card
sorting task than others the first time it is performed?
Chi-square
value
Degrees of
freedom
Significance
level
χ2 (1) = 19.74, p < .001
Chi-Square Tests
Pears on Chi-Square
Continuity Correctiona
Likelihood Ratio
Fisher's Exact Test
Linear-by-Linear
Ass ociation
N of Valid Cas es
Value
19.740b
18.193
20.887
19.631
df
1
1
1
1
Asymp. Sig.
(2-s ided)
.000
.000
.000
Exact Sig.
(2-s ided)
Exact Sig.
(1-s ided)
.000
.000
.000
180
a. Computed only for a 2x2 table
b. 0 cells (.0%) have expected count les s than 5. The minimum expected count is
20.50.
WASON CARD SORT
RESULTS
Q1: Is performance better on some versions of the Wason card
sorting task than others the first time it is performed?
You will need to graphically
represent your results, too
Don’t forget to give your figure a
number and a title
When you refer to the figure in the
main text, make sure you give the
exact descriptive statistics
Note: no error bars, because this is
qualitative data
WASON CARD SORT
RESULTS
Q2: Is performance on the abstract task affected if it follows a
concrete scenario?
ABSTRACT (1st) vs.
ABSTRACT (2nd)
If we are interested in comparing
categorises of responses, then the
chi-square test is appropriate.
In SPSS, the chi-square test is
hidden away underneath
descriptive statistics > crosstabs.
Let’s go there now.
WASON CARD SORT
RESULTS
Q2: Is performance on the abstract task affected if it follows a
concrete scenario?
ABSTRACT (1st) vs.
ABSTRACT (2nd)
In order to build a chi-square table,
we need to put our various
categories into rows and columns.
Let’s put first_pr as a row and
correc_a as a column. This will show
us the frequency distributions.
Under statistics, we also need to
make sure the chi-square test is
performed, so tick that.
Under cells, also make sure that both
observed and expected are clicked.
WASON CARD SORT
RESULTS
Q2: Is performance on the abstract task affected if it follows a
concrete scenario?
ABSTRACT (1st) vs.
ABSTRACT (2nd)
CORRECTA * FIRST_PR Crosstabulation
FIRST_PR
1
CORRECTA
right
wrong
Total
Count
Expected Count
Count
Expected Count
Count
Expected Count
2
8
9.5
82
80.5
90
90.0
11
9.5
79
80.5
90
90.0
Total
19
19.0
161
161.0
180
180.0
The second table (after case
processing summary) confirms
our 180 observations and displays
the frequency distribution of right
and wrong responses for the two
kinds of Wason card sort test.
Chi-Square Tests
Pears on Chi-Square
Continuity Correctiona
Likelihood Ratio
Fisher's Exact Test
Linear-by-Linear
Ass ociation
N of Valid Cas es
Value
.530 b
.235
.532
.527
df
1
1
1
1
Asymp. Sig.
(2-s ided)
.467
.628
.466
Exact Sig.
(2-s ided)
Exact Sig.
(1-s ided)
.629
.314
The third table provides us with
the chi-square value, which may
be reported as:
.468
180
a. Computed only for a 2x2 table
b. 0 cells (.0%) have expected count les s than 5. The minimum expected count is
9.50.
χ2 (1) = 0.53, p = .47
WASON CARD SORT
RESULTS
Q2: Is performance on the abstract task affected if it follows a
concrete scenario?
Once again, you will need to graphically represent your results
Don’t forget to give your figure a number and a title
When you refer to the figure in the main text, make sure you give the exact
descriptive statistics
Note: no error bars, because this is qualitative data
If you have trouble, refer back to the Excel graph-making guides on
Graham’s webpage, or ask a tutor for help
WASON CARD SORT
DISCUSSION
GET TOGETHER IN GROUPS OF THREE OR FOUR AND REFLECT
ON TODAY’S EXPERIENCE USING THE FOLLOWING QUESTIONS
Why have I done this
particular statistical test?
What do the data actually
tell me with respect to my
experimental hypotheses?
What implications do the
data have for the studies
outlined in the intro?
SAMPLING DISTRIBUTIONS
LEARNING OUTCOMES
BY THE END OF THE SESSION, YOU SHOULD BE ABLE TO:
Use SPSS to test the first experimental hypothesis of the Wason
card sorting experiment and produce a related graph.
Use SPSS to test the second experimental hypothesis of the
Wason card sorting experiment and produce a related graph.
Use the graphs to interpret your chi-square findings.
Make a start on writing up your RESULTS and DISCUSSION
sections for your lab report.