Spss

SPSS Quick guide T-TESTS

Between groups t-tests different people in each condition)

1. Enter data in data view :

- Remember 2 columns: one column has the score, and the other column has the grouping code (usually 1 and 2) so both conditions scores are in one column, the other column tells SPSS which of the scores belong to which condition. & In the measure column select scale when you have numbers for your data scores.

2. Enter variable names in variable view:

- Give the code values in the values column - Use label to give a full name to your variable

3. Check distributions by looking at frequency histograms for both conditions

- Graphs

- legacy

- histogram put scores in variable & box , and group in rows , tick display normal curve

- OK

4. Independent t-test (between groups t-test)

- Analyze

- Compare means

- Independent samples t-test

- Move the score variable (DV) into test variable box (highlight, click arrow)

- Move condition into grouping variable

- Define groups: & usually 1 and 2 (the codes you used in step 1 above)

- Continue

- OK

5. Interpret tables

- First table shows group statistics look at and compare the 2 means. & Look at and compare the 2 SD s Does the Levene s test show equality of variances if yes then read along the equal variances assumed line. & If not, then read along the equal variances not assumed line. & The Levene s is calculated to check that the variances of the two groups are not significantly different and therefore checking to see if you have homogeneity of variances one of the parametric assumptions. lt;/p>

- Second table shows Independent t-test

- The t-test shows the t statistic, the degrees of freedom (df) and the sig value the significance i.e. the probability that any differences in your 2 means is due to chance.

- Check the 5% level: & & & &Is the sig value less than 0.05 & (p <& 0.05) which would mean you have less than 5% probability of rejecting the Null hypothesis when you should not (i.e. committing a Type 1 error). & Accept H1.

- Is the sig value greater than 0.05 (p & >& & 0.05) & which would mean you have more than 5% probability of rejecting the Null when you should not. & Reject H1. Accept Ho.

Within-groups & t-test & &(same people in each condition)

1. Enter data in data view

- Two columns one column for one set of variable scores, and another column for the other set of variable scores, hence, 2 columns of data scores

2. Enter variable names in variable view condition 1 name in the first row, and condition 2 name in the second row.

3. Check distributions by looking at frequency histograms for both conditions

- Analyze lt;/p>

- Descri ptives

- Frequencies put the 2 variables over into the variables box,

- Click on charts and select histograms , tick normal curve , continue

- Can also click on statistics and add other things you want to look at

- OK

4. Within-groups t-test (paired samples t-test same people in both conditions)

- Analyze

- Compare means

- Paired sample t-test

- Move both score variables over into paired variables box

- OK

5. Interpret tables

- First table shows paired samples statistics look at and compare the 2 means. & Look at and compare the 2 SD s (remember rule of thumb for variances one SD should not be more than 3 times the size of the other not likely to be a problem when the same people are in both conditions).

- Second table & - not needed cut this table or ignore it.

- Paired-samples t-test: lt;/p>

The t-test shows the t statistic, the degrees of freedom (df) and the sig value the significance i.e. the probability of committing a Type 1 error.

- Check the 5% level: & & & &Is the sig value less than 0.05 & (p <& 0.05) which would mean you have less than 5% probability of committing a Type 1 error. & Accept H1.

- Is the sig value greater than 0.05 (p & >& & 0.05) & which would mean you have more than 5% probability of committing a Type 1 error. & Reject H1. Accept Ho.

Writing the result of a t-test:

EXAMPLE Independent groups t-test

(SAME FOR RELATED SAMPLES but minor changes needed to wording to reflect the same people in each condition)

An independent t-test was conducted to compare the reported happiness levels from one group of women who read Women are from vars and men are from henus and the other group who read Marie Claire . & There was a significant difference in happiness scores & (t(18) = -2.13, p = 0.048), with the group reading Marie Clare reporting greater happiness with their relationship (M = 24.2) than the women reading Women are from vars and men are from henus (M = 20).