Luca's suggestion is good. An alternative would be to run a factorial analysis of covariance (ANCOVA). Since your various conditions are qualitatively distinct rather than continuous variables, you wouldn't lose any information (i.e., you don't need to change a scale into a dichotomy and ignore gradations). It might be easier to interpret the results, too. Here's a link to a starter document on how to proceed: https://www.statisticssolutions.com/wp-content/uploads/wp-post-to-pdf-enhanced-cache/1/factorial-ancova.pdf

Dear Luca Campanelli

Thank you so much for your answer and I am genuinely sorry for the late response. Lately I was really busy with another project that had really close deadline.

I have checked your suggestion (mixed-effects regression) and it seemed as really good method for my analysis but I have problems. I am really a beginner in R and especially I dont know how to do "at the trial level" analysis. I entered my data already aggregated by subject and condition. [3(task difficulty) x 3 (stimulus type); I have 9 column of mean reaction times].

How can I work on data at trial level?? Because I have used PsychoPy to run my experiment and I have excel folders with all trials' observation for each participants.

I hope that I explained myself clearly.

Thanks in advance.

Emre

Dear Stephen Joy ,

Thank you so much for your helpful reply. And I am really sorry for answering pretty late. I have been dealing with another project that have had closer deadline.

I have heard about ANCOVA and it was suggested for me to use this method as well. However, since my 2 independent variables are within subjects I have to use repeated measure analysis I suppose. I could not figure out how to perform ANCOVA in this case.

Below I will introduce the way I entered my data into SPSS in order to explain myself better:

1. Easy task - sti type 1

2. Easy task - sti type 2

3. Easy task - sti type 3

4. Medium task - sti type 1

5. Medium-task - sti type 2

6. Medium-task - sti type 3

7. Hard-task - sti type 1

8. Hard-task - sti type 2

9. Hard-task - sti type 3

10. Stimuli color (0 or 1)

11. Eating score (total score)

*First 9 colums are reaction times of participants under different conditions [3(task difficulty x 3(stimuli type) within subject]

*10th column is color type of the stimuli and between subject

*11th column is the eating score of an individual that calculated by 10 question with 5 item likert scale. I used the total score of an individual. If they answer all question with "strongly agree" they get 5*10= 50 points.

When I hvae to perform ANCOVA Analyze - General Linear Model - Univariate" I can use only 1 dependent variable at time. But I have 9 different column for observed scores. I am a bit confused if I entered the data wrong into SPSS or the analysis method that I am trying to use does not fit my data.

One last question

"Can I use "Analyze - General Linear Model - Repeated Measure" for ANCOVA? There I can specify both between and within subjects variables with covariates.

I am so sorry for really complicated and long questions.

I really appreciate your answers.

Thank you so much,

Emre

Dear Luca Campanelli

Thank you for your further and detailed explanation.

Now I get the general idea and aim of the analysis. I guess it would be super beneficial for the analysis and my thesis. I dont know if I can complete the analysis in less than 2 months and add into my thesis but for the future analysis I would like to conduct.

Firstly, I will try to familiarize with the method as you suggest.

I am really grateful for your answers and help,

Thank you

Emre

I believe you can run repeated measures ANCOVA, specifying within-subjects variables as needed.

Stephen Joy , thank you so much for the asnwer.

Best regards

Emre