Thursday, May 3, 2018

Assalamualaikum W.B.T!

The purpose of this blog was created is because we want to show how we run the IBM SPSS.

We had been assigned to run an independent samples t-test as our project for subject PSYC 1998 (Psychological Statistics).

First of all, we need to choose a topic and we had agreed to measure body self esteem and eating attitude between male and female among IIUM Undergraduates students as our title for the project.

We decided to use The Body Esteem Scale-Revised by Frost, Franzoi, Oswald and Shields (2018) and Eating Attitudes Test (EAT-26) by Gardner and Garfinkel (1982) as our scale.

Next later we used survey method to spread our survey among the IIUM Undergraduate students.

After completed collecting the data we need to analyze three things using IBM SPSS which are

a) Descriptive statistical analysis
b) The Independent samples T-Test
c)  The Reliability analysis- Cronbach's Alpha of the two scale that we used.


Below is the step on how we completed the tasks above using the IBM SPSS.




1) Firstly, we prepare a coding strategy in order to tally the original data sheet with the SPSS data file.

To perform this step, firstly we open the excel file that already contain our raw data.






Next, we change the value by clicking button "ctrl" and "H" at the same time and there is a dialog box will appear and then we give a texts label for category codes for the variable. We key-in female (the variable) on the "find what"  put the value 1 on the "replace with" like the picture below.



Then the variable will change into the value like the picture below.





2) The second step is we create data file and key-in the data.

 First, we double-clicked the IBM SPSS icon on the desktop and the window like the picture below will appear.


  
Next, click file > open > data.
                                    


After that, there will be a window appear like on the picture below. In order to open the existing data file from excel, change the data type to the excel first. Then click the excel file that containing the raw data and click open.


Then click on the window that appear.



Then we will be present with the data like in the picture below.


After that, there will be three things that we need to fix which are the label, variable value and variable measure. In order to change the label  just give a long description of the variable while for the variable value click on the value cell and then click the button that have three small dots on it. This button will open the value label on the dialog box. For the value, key-in with number "1.00" while for the label type in Female and then click "OK". Repeat the same steps for all of the variable.


Next, change the variable measure into the appropriate level of measurement and then chose scale for the item of the scale and choose nominal for the gender, kuliyyah and level of study.


After complete the steps above, we will be able to see the completed data file.

The data view will look like this.


Meanwhile the variable view will look like in the picture below.


3) Next, we prepare the codebook.

To prepare the codebook we click on analyze > reports > codebook.


Then, we will be presented with the dialog box and for the variable tab move the variable that we want to the codebook variable box using the arrow ↪.


  And then click the output tab and untick all except for the label, value label and missing value.                            

Meanwhile for the statistic tab, tick all option and then click "OK"


After that we will be presented with the codebook output.


4) Next, we do screening, cleaning and preparing the data

This step is to check if there are any data entry error via frequency table.

First we click analyze > descriptive statistics > frequencies.


After that the frequency dialog box will appear. Move the variable we need to the variable box using this arrow ↪ and click "OK". If there is any incorrect value go back to the data view and correct the value again.


Next, we compute our variable based on the existing information in the data. In this step, we compute the total score for each participant for scale 1. 

To compute click on the transform > compute variable


After that, we will be presented with the new window below. The steps we use to compute are like below :
  1. Target variance. We change the target variances to "totbody".
  2. Function group: click "All".
  3. Function and special variables: Click "Sum".
  4. Click this button arrow ⤴.
  5. Type and label: Move the first and the last variable we want to compute to the numeric expression. We move body as it is our first variable and later move coordination as it is our last variable.  (SUM(Body to Coordination))



5) After that we recheck our codebook and also check the statistical assumption.

We recheck the codebook by looking at codebook output. We focused on the missing value part. We want to see  if we have any missing value.




After that, we check the statistical assumption.


So for the first step, we click on analyze > descriptive statistics > explore.


Next, we will be presented with a new window. Move the variable data that we want to the dependent list using this arrow ↪.



After that click the plots tab and the new window will appear. Tick "None" for boxplot, and tick everything for descriptive box except for stem-and-leaf  and also tick the box "Normality plots with tests". And click "Continue".


And the normality output will appear.



6) Next, we conduct the appropriate descriptive analysis.

The first step to perform descriptive analysis is to click analyze > Descriptive statistics >Descriptive.



The descriptive box will appear and we move the variable to the variables. Then click option and tick "Mean, Sum, Std, Deviation, Variances, Min, Max, Kurtosis, Skewness, Variable list " and click "Continue".




Then the descriptive output will appear.




7) After that we conduct Independent  samples T- test.

To conduct the independent samplest- test first click analyze > compare means > independent sample t- test.




Then we move the variable to the tests variable and for grouping variable we move the gender and define the group.



Then we define the group by using specific values. For group 1 "1.00" and for group 2 "2.00".






Then we click the option button and make sure the confidence interval percentage is 95%.



Then the independent samples t- test output will appear.









8) After that we report the result in APA format. In order to report the result to APA format we need to change the table that we got from the output into APA format.

First we right click > edit content > in viewer.



After it become like in the picture below right click it again and click table look.


After that the new window will pop out and we choose compact academic and then we click edit look button.


Then we are presented with the table properties window.

On the border tab we click "Top inner frame" and for the style we choose the thin line style.



After that we click "Cell formats" tab and the tab like in the picture below will appear.

In the area section click "Title" while for the texts : change the font to "Times new roman with size 12".
After that unbold it and click the italic button.(The italic is just for the title).

Repeat the action for all the item in the area.

After finished it click "OK".



After that we will get the table in the APA format.

We repeat this steps to convert other tables that we need to report into APA format table.







                                                                                           RESULT SECTION



In the result section we will show how we report the result of the

  • Descriptive statistical analysis
  • Independent samples t-test
  • The Cronbach's alpha (the reliability of test)
in the APA format.


Report of the result for descriptive statistics



The body esteem scale (M = 84.2750, SD= 15.46375). Meanwhile the eating attitude test (M = 13.2750, SD= 10.44150). The median of the body esteem scale was 87 while the median of eating attitude test was 10. 89 occur the most therefore 89 is the mode for the body esteem scale while 3 occur the most, therefore 3 is the mode for eating attitude test.



Report of normality assumption





The tests used was the Shapiro-Wilk test and the sample sized for body esteem scale shown is 40 and significant value of Shapiro-Wilk p > 0.05. Therefore the data is approximately normally distributed.

The tests used was the Shapiro-Wilk test and the sample sized for eating attitudes test scale shown is 40 and significant value of Shapiro-Wilk p < 0.05. Therefore the data was significantly deviated from the normal distribution.




Report of the result for Independent samples t- test



A independent-sample t-test was conducted to examine whether there were significant difference between female's body esteem score and male's body esteem score. Results showed that is a significant difference in the body esteem score between female ( M = 78.55, SD = 13.39864) and male ( M= 90.00, SD = 15.56650), t(38) = -2.493, p < .05 ( two-tailed),  Cohen's d0.7884. Therefore we can conclude that there were significant difference between female's body esteem score and male's body esteem score.

A independent-sample t-test was conducted to examine whether there were significant difference between female's eating attitude score and male's eating attitude score.Result showed that there is a significant difference in the eating attitude score between female ( M =  15.90, SD = 9.67308 ).and male ( M = 10.65, SD = 10.75701), t(38) = 1.623, p > .05 (two-tailed). Therefore there was no significant difference between  female's eating attitude score and male's eating attitude score.








Report of reliability test

Scale 1:




The body esteem scale was used to measure to the body esteem of  IIUM undergraduate students. A total of 40 undergraduate students at International Islamic University Malaysia answered this scale, which consisted of 24 items. The Cronbach's alpha for these 24 items was 0.932, indicating an excellent reliability.


Scale 2:




The eating attitude test (EAT-26) was used to measure to help to determine whether IIUM undergraduate students have eating disorders. A total of 40 undergraduate students at International Islamic University Malaysia answered this scale, which consisted of 26 items. The Cronbach 's alpha for these 26 items was 0.897, indicating a good reliability.




REFERENCES


Cronbach, L. J. (1951) Coefficient alpha and the internal structure of tests. Psychometrika, 16(3),  297-334.


Frost, K.A., Franzoi, S.L., Oswald, D.L., & Shields, S.A. (2018). Revising the Body Esteem Scale with a U.S. college student sample: Evaluation, validation, and uses for the BES-R.   Sex Roles, 78, 1-17. doi: 10.1007/s11199-017-0776-5.

Garner, D.M., Olmsted, M.P., Bohr, Y., and Garfinkel, P.E. (1982). The Eating Attitudes Test: Psychometric features and clinical correlates. Psychological Medicine, 12, 871-878.



-----------------------------------------------------------------------------------------------------------------------------------------------------------------






                                                                              Group members:
                                                                  1)Nurril Tasnim ABd Ghaffar (1626406)
                                                                  2) Wan Nurul Hanna Saidah Reduan (1625232)
                                                                  3) Nursabrina Zaidi (1629756)
                                                                  4)Aisyah Hamizah Abd Rahman (1624408)
                                                                  5) Nursyahidatul Syahira Mazlan (1620932)

Assalamualaikum W.B.T! The purpose of this blog was created is because we want to show how we run the IBM SPSS. We had been assigned to ...