# Thread: Post-test only design & ANCOVA/ANOVA

1. ## Post-test only design & ANCOVA/ANOVA

Hello,

I need some help regarding the post-test only design.

If, in a post-test only quasi-experimental study having 3x2 factorial design, three groups of students are involved & each group is supposed to be taught by a different teaching method.

In this design, there are:

1. two independent variables:

(a) teaching methods (3 levels/groups) (b) gender

2. three dependent variables:

(a) achievement (3 sublevels/subscales)

(b) attitude towards Science (measured by a likert scale consisting of 7 subscales)

(c) attitude towards use of technology in Science teaching (measured by a likert scale consisting of 5 subscales)

3. three covariates:

(a) IQ
(c) pre-attitude scores towards teaching subject

Pre-test achievement can not be measured as the students are supposed to be taught certain topics which are entirely new to them & pre-attitude towards use of technology in Science teaching also can not be measured as students have so far not experienced the use of any kind of technology by their Science or other subjects teachers in the classroom.

Is it appropriate to conduct ANCOVA on each dependent variable by adjusting for covariates? Or, will ANOVA be sufficient enough (as groups are found to be equivalent on all covariates)?

2. ## Re: Post-test only design & ANCOVA/ANOVA

The post-test option would work best if the methods were randomly assigned to the students. This would assume that their variability was proportionate between experiemental groups. Were you able to do this? What is your sample size? What covariance would you be trying to control for?

3. ## The Following User Says Thank You to hlsmith For This Useful Post:

indicastats (10-26-2012)

4. ## Re: Post-test only design & ANCOVA/ANOVA

Originally Posted by hlsmith
The post-test option would work best if the methods were randomly assigned to the students. This would assume that their variability was proportionate between experiemental groups. Were you able to do this? What is your sample size? What covariance would you be trying to control for?
3 intact groups are taken & teaching methods are assigned to them randomly.

Total sample size is 210 (70 students in each group).

3 groups are not significantly different from each other as far as their scores on covariates (as mentioned above) are concerned. Shall I go for ANOVA or shall I still have to adjust for the covariates?

6. ## Re: Post-test only design & ANCOVA/ANOVA

With what you have described I may proceed with ANOVA/linear regression controlling for the covariates and also examining for collinearity.

7. ## The Following User Says Thank You to hlsmith For This Useful Post:

indicastats (10-26-2012)

8. ## Re: Post-test only design & ANCOVA/ANOVA

Originally Posted by hlsmith
With what you have described I may proceed with ANOVA/linear regression controlling for the covariates and also examining for collinearity.
So you mean to say that ANOVA can be done in case of post-test only QUASI-EXPERIMENTAL design when groups are found to be equivalent on covariates.

I have read in books about using ANOVA in case of post-test only design when subjects are assigned to groups randomly as well as teaching methods are randomly assigned to groups. But since in my case only random assignment of teaching methods to intact groups was possible, I was a little worried about using ANOVA.

Can you please elaborate a little more on linear regression? How to proceed with linear regression in this design? Why?

9. ## Re: Post-test only design & ANCOVA/ANOVA

ANOVA is a special case of linear regression. I recommended multiple linear regression because it is typically used when you have multiple independent and control variables that will be in the model. Are you planning on running three models for the three dependent measures of interest?

10. ## The Following User Says Thank You to hlsmith For This Useful Post:

indicastats (10-26-2012)

11. ## Re: Post-test only design & ANCOVA/ANOVA

Originally Posted by hlsmith
ANOVA is a special case of linear regression. I recommended multiple linear regression because it is typically used when you have multiple independent and control variables that will be in the model. Are you planning on running three models for the three dependent measures of interest?

Sorry I did not understand : running three models for the three dependent measures of interest

but how to do it in SPSS.

12. ## Re: Post-test only design & ANCOVA/ANOVA

No idea how to do that in SPSS. I was going back to you original post where you stated that you have 3 dependent variables, so I did not know if you were planning on repeating your procedures 3 times, since you have 3 things to predict (3 dependent variables).

After quickly reading over this thread, I believe the literature you reviewed about randomizing a second variable in ANOVA may be in regards to controlling for it via randomization. If you did not randomize the second variable, you can always put it in your model (say in multivariable linear regression) to see if it has an effect.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts