Aim/ Overview of the practical- What are the Application of Chi-Square
Test and apply other functions on them also
To know how the data is entered in SPSS.
To know how to carry out the Chi-square test for nominal and ordinal
data.
To understand the fact that whether the variables are dependent or if
they have any association between them.
Interpreting the results from the output.
Task to be done-
a) Steps of the practical- Take the data and analyze them using Analyze option
1. First of all, in the data view tab, click on analyse, then choose descriptive
statistics and then open Crosstabs.
2. Now, choose the Variables that you want to analyze and enter them in the
rows and columns boxes.
3. Click on Exact and then select Asymptotic only.
4. Click on Statistics and select Chi-Square.
5. Select the tests under the nominal and ordinal heading. (If both the
variables have nominal data then you can select all or any test under Nominal
and same conditions apply to ordinal into ordinal variables that we can choose
any test under the ordinal head. In case one variable is ordinal and one is
nominal, then select Phi and Cramer’s V only.) For our tests we will work on
Phi and Cramer’s V.
6. Click on continue.
7. Click on cells and select observed and expected, row and total or column
and total.
8. Click on continue and then Format, select ascending and descending
depending on choice.
9. Click on continue
10. Click OK.
11. The output screen will appear
a) Output (Write in brief) -
Explanation for Expected Counts
Checking the assumption whether the expected count in the output is less or more than 5. If it is
less than 5 then we cannot perform the Chi-Square on that because an assumption has been
violated. Thus, despite having the results of the test in the output, if we draw any conclusion
based on them, or take any decision, then the chances of failure will be high. In case all the
expected counts are more than 5, then the assumption will be qualified.
Our Output
Since in our output, we have expected count values less than 5 in case of all the tables, i.e.,
gender and help required table, gender and frequency table, SBI Account and visiting week table,
age and ID related table and type of transaction and type of bank account table, thus the
assumption will not be qualified.
Explanation for the Chi-Square Tests Table
After checking the expected count we come down to the Chi-Square tests table in the output
screen and check the value for the Pearson Chi-Square. The value indicates the strength of
association, but from this value it cannot be analyzed whether the strength of association is low,
medium or high.
For checking the Strength of Association
For checking the significant association between the variables, we check the Phi value. If Phi
Value is less than or equal to 0.05, then there is significant association and thus the null
hypothesis is rejected.
For checking the strength of association we check the values in the value column of the
symmetric measures. The values should lie between 0 to1. If the value is closer to 0, then we can
say that the strength of association is poor and vice versa.
The Generated Output
In case of gender of the respondent * if the help of the general manager taken, the null
hypothesis is rejected as the Phi value is less than 0.05, forming a significant association,
but the strength of association is week.
In case of gender of the respondent * frequency of visits to the bank, the null hypothesis
is accepted as the Phi value is more than 0.05, forming no significant association.
In case of whether the respondent has an SBI Account * in which week does the visitor
visits, the null hypothesis is accepted as the Phi value is more than 0.05, forming no
significant association.
In case of type of transaction carried out * type of bank account, the null hypothesis is
accepted as the Phi value is more than 0.05, forming no significant association.
In case of id of the respondent * age of the respondent, the approximate significance
value is less than 0.05, forming a significant association, but the strength of association is
week.
b) Images of the steps and output-
Variable View is Where we can additional information of our data
Data view is where we can inspect our actual data
Images of the steps and output-
In case of nominal variables, following are the common steps:
Step 1:
STEP 2:
STEP 3:
STEP 4:
STEP 5:
STEP 6:
STEP 7: The output will appear
CROSSTABS
/TABLES=gender BY sbiacc
/FORMAT=AVALUE TABLES
/STATISTICS=CHISQ PHI
/CELLS=COUNT ROW COLUMN TOTAL
/COUNT ROUND CELL.
Crosstabs
Notes
Output Created 06-Sep-2020 23:42:34
Comments
Input Data C:\Users\VK\Desktop\[Link]
Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data 25
File
Missing Value Handling Definition of Missing User-defined missing values are treated
as missing.
Cases Used Statistics for each table are based on
all the cases with valid data in the
specified range(s) for all variables in
each table.
Syntax CROSSTABS
/TABLES=gender BY sbiacc
/FORMAT=AVALUE TABLES
/STATISTICS=CHISQ PHI
/CELLS=COUNT ROW COLUMN
TOTAL
/COUNT ROUND CELL.
Resources Processor Time 00:00:00.000
Elapsed Time 00:00:00.014
Dimensions Requested 2
Cells Available 174762
[DataSet1] C:\Users\VK\Desktop\[Link]
Case Processing Summary
Cases
Valid Missing Total
N Percent N Percent N Percent
gender of the customer * do 25 100.0% 0 .0% 25 100.0%
they have sbi account
gender of the customer * do they have sbi account Crosstabulation
do they have sbi account
NO
gender of the customer Count 2 0
% within gender of the 100.0% .0%
customer
% within do they have sbi 100.0% .0%
account
% of Total 8.0% .0%
FEMALE Count 0 2
% within gender of the .0% 20.0%
customer
% within do they have sbi .0% 40.0%
account
% of Total .0% 8.0%
MALE Count 0 3
% within gender of the .0% 23.1%
customer
% within do they have sbi .0% 60.0%
account
% of Total .0% 12.0%
Total Count 2 5
% within gender of the 8.0% 20.0%
customer
% within do they have sbi 100.0% 100.0%
account
% of Total 8.0% 20.0%
gender of the customer * do they have sbi account Crosstabulation
do they have sbi
account
YES Total
gender of the customer Count 0 2
% within gender of the .0% 100.0%
customer
% within do they have sbi .0% 8.0%
account
% of Total .0% 8.0%
FEMALE Count 8 10
% within gender of the 80.0% 100.0%
customer
% within do they have sbi 44.4% 40.0%
account
% of Total 32.0% 40.0%
MALE Count 10 13
% within gender of the 76.9% 100.0%
customer
% within do they have sbi 55.6% 52.0%
account
% of Total 40.0% 52.0%
Total Count 18 25
% within gender of the 72.0% 100.0%
customer
% within do they have sbi 100.0% 100.0%
account
% of Total 72.0% 100.0%
Chi-Square Tests
Asymp. Sig. (2-
Value df sided)
Pearson Chi-Square 25.034a 4 .000
Likelihood Ratio 13.970 4 .007
N of Valid Cases 25
a. 7 cells (77.8%) have expected count less than 5. The minimum
expected count is .16.
Symmetric Measures
Value Approx. Sig.
Nominal by Nominal Phi 1.001 .000
Cramer's V .708 .000
N of Valid Cases 25
CROSSTABS
/TABLES=acctype BY tran_typ
/FORMAT=AVALUE TABLES
/STATISTICS=CHISQ PHI
/CELLS=COUNT ROW COLUMN TOTAL
/COUNT ROUND CELL.
Crosstabs
Notes
Output Created 06-Sep-2020 23:48:50
Comments
Input Data C:\Users\VK\Desktop\[Link]
Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data 25
File
Missing Value Handling Definition of Missing User-defined missing values are treated
as missing.
Cases Used Statistics for each table are based on
all the cases with valid data in the
specified range(s) for all variables in
each table.
Syntax CROSSTABS
/TABLES=acctype BY tran_typ
/FORMAT=AVALUE TABLES
/STATISTICS=CHISQ PHI
/CELLS=COUNT ROW COLUMN
TOTAL
/COUNT ROUND CELL.
Resources Processor Time 00:00:00.000
Elapsed Time 00:00:00.006
Dimensions Requested 2
Cells Available 120988
[DataSet1] C:\Users\VK\Desktop\[Link]
Case Processing Summary
Cases
Valid Missing Total
N Percent N Percent N Percent
type of account * transaction 25 100.0% 0 .0% 25 100.0%
type
type of account * transaction type Crosstabulation
transaction type
demand draft deposits
type of account Count 2 3 0
% within type of account 28.6% 42.9% .0%
% within transaction type 100.0% 60.0% .0%
% of Total 8.0% 12.0% .0%
CURRENT Count 0 1 1
% within type of account .0% 50.0% 50.0%
% within transaction type .0% 20.0% 8.3%
% of Total .0% 4.0% 4.0%
LOAN Count 0 0 1
% within type of account .0% .0% 100.0%
% within transaction type .0% .0% 8.3%
% of Total .0% .0% 4.0%
OTHER Count 0 0 1
% within type of account .0% .0% 100.0%
% within transaction type .0% .0% 8.3%
% of Total .0% .0% 4.0%
PPF Count 0 0 1
% within type of account .0% .0% 100.0%
% within transaction type .0% .0% 8.3%
% of Total .0% .0% 4.0%
SAVING Count 0 1 8
% within type of account .0% 7.7% 61.5%
% within transaction type .0% 20.0% 66.7%
% of Total .0% 4.0% 32.0%
Total Count 2 5 12
% within type of account 8.0% 20.0% 48.0%
% within transaction type 100.0% 100.0% 100.0%
% of Total 8.0% 20.0% 48.0%
type of account * transaction type Crosstabulation
transaction type
withdrawal Total
type of account Count 2 7
% within type of account 28.6% 100.0%
% within transaction type 33.3% 28.0%
% of Total 8.0% 28.0%
CURRENT Count 0 2
% within type of account .0% 100.0%
% within transaction type .0% 8.0%
% of Total .0% 8.0%
LOAN Count 0 1
% within type of account .0% 100.0%
% within transaction type .0% 4.0%
% of Total .0% 4.0%
OTHER Count 0 1
% within type of account .0% 100.0%
% within transaction type .0% 4.0%
% of Total .0% 4.0%
PPF Count 0 1
% within type of account .0% 100.0%
% within transaction type .0% 4.0%
% of Total .0% 4.0%
SAVING Count 4 13
% within type of account 30.8% 100.0%
% within transaction type 66.7% 52.0%
% of Total 16.0% 52.0%
Total Count 6 25
% within type of account 24.0% 100.0%
% within transaction type 100.0% 100.0%
% of Total 24.0% 100.0%
Chi-Square Tests
Asymp. Sig. (2-
Value df sided)
Pearson Chi-Square 16.513a 15 .349
Likelihood Ratio 20.732 15 .146
N of Valid Cases 25
a. 23 cells (95.8%) have expected count less than 5. The minimum
expected count is .08.
Symmetric Measures
Value Approx. Sig.
Nominal by Nominal Phi .813 .349
Cramer's V .469 .349
N of Valid Cases 25
CROSSTABS
/TABLES=visit BY per_visi
/FORMAT=AVALUE TABLES
/STATISTICS=CHISQ PHI
/CELLS=COUNT ROW COLUMN TOTAL
/COUNT ROUND CELL.
Crosstabs
Notes
Output Created 06-Sep-2020 23:49:32
Comments
Input Data C:\Users\VK\Desktop\[Link]
Active Dataset DataSet1
Filter <none>
Weight <none>
Split File <none>
N of Rows in Working Data 25
File
Missing Value Handling Definition of Missing User-defined missing values are treated
as missing.
Cases Used Statistics for each table are based on
all the cases with valid data in the
specified range(s) for all variables in
each table.
Syntax CROSSTABS
/TABLES=visit BY per_visi
/FORMAT=AVALUE TABLES
/STATISTICS=CHISQ PHI
/CELLS=COUNT ROW COLUMN
TOTAL
/COUNT ROUND CELL.
Resources Processor Time 00:00:00.015
Elapsed Time 00:00:00.072
Dimensions Requested 2
Cells Available 98303
[DataSet1] C:\Users\VK\Desktop\[Link]
Case Processing Summary
Cases
Valid Missing Total
N Percent N Percent N Percent
do they visit * per visit in a 25 100.0% 0 .0% 25 100.0%
week
do they visit * per visit in a week Crosstabulation
per visit in a week
1st week 2nd week
do they visit Count 2 4 1
% within do they visit 28.6% 57.1% 14.3%
% within per visit in a week 100.0% 28.6% 25.0%
% of Total 8.0% 16.0% 4.0%
less than one week Count 0 1 1
% within do they visit .0% 33.3% 33.3%
% within per visit in a week .0% 7.1% 25.0%
% of Total .0% 4.0% 4.0%
more than three months Count 0 0 0
% within do they visit .0% .0% .0%
% within per visit in a week .0% .0% .0%
% of Total .0% .0% .0%
one month to three months Count 0 2 2
% within do they visit .0% 40.0% 40.0%
% within per visit in a week .0% 14.3% 50.0%
% of Total .0% 8.0% 8.0%
one week to one month Count 0 7 0
% within do they visit .0% 77.8% .0%
% within per visit in a week .0% 50.0% .0%
% of Total .0% 28.0% .0%
Total Count 2 14 4
% within do they visit 8.0% 56.0% 16.0%
% within per visit in a week 100.0% 100.0% 100.0%
% of Total 8.0% 56.0% 16.0%
do they visit * per visit in a week Crosstabulation
per visit in a week
3rd week 4th week Total
do they visit Count 0 0 7
% within do they visit .0% .0% 100.0%
% within per visit in a week .0% .0% 28.0%
% of Total .0% .0% 28.0%
less than one week Count 1 0 3
% within do they visit 33.3% .0% 100.0%
% within per visit in a week 33.3% .0% 12.0%
% of Total 4.0% .0% 12.0%
more than three months Count 1 0 1
% within do they visit 100.0% .0% 100.0%
% within per visit in a week 33.3% .0% 4.0%
% of Total 4.0% .0% 4.0%
one month to three months Count 0 1 5
% within do they visit .0% 20.0% 100.0%
% within per visit in a week .0% 50.0% 20.0%
% of Total .0% 4.0% 20.0%
one week to one month Count 1 1 9
% within do they visit 11.1% 11.1% 100.0%
% within per visit in a week 33.3% 50.0% 36.0%
% of Total 4.0% 4.0% 36.0%
Total Count 3 2 25
% within do they visit 12.0% 8.0% 100.0%
% within per visit in a week 100.0% 100.0% 100.0%
% of Total 12.0% 8.0% 100.0%
Chi-Square Tests
Asymp. Sig. (2-
Value df sided)
Pearson Chi-Square 21.873a 16 .147
Likelihood Ratio 20.995 16 .179
N of Valid Cases 25
a. 24 cells (96.0%) have expected count less than 5. The minimum
expected count is .08.
Symmetric Measures
Value Approx. Sig.
Nominal by Nominal Phi .935 .147
Cramer's V .468 .147
N of Valid Cases 25
LEARNING OUTCOMES
How to enter data in variable and nominal view.
How to carry out chi Square test of association.
How to interpret the result from output.