I have two experiments with very similar procedures and materials. I would like to compare one EG from the first experiment with one EG from the second experiment. Samples in both studies are also very similar in terms of size (around 30 people) and characteristics (students from the same faculty). Is there any way to do this?
You can compare two groups error in the experiment results. Comapre two group results have a results distribution for the accepted value by a method with bias. The results distribution from two groups that could have a (significant) bias.
Thank you very much Mr. Chen. Do you know any method to do this with SPSS? If you could indicate any interesting reference to learn how to do this comparison, it will be much appreciated!
Dear Carlos,
I'm not an expert, but I would consider this a 2-level model, with the effects of interest nested in the experiments, in which case you can analyze it in spss with mixel>linear, and ask for the estimated marginal means as output. You can then compare based on confidence interval overlap, or significance testing. Good luck!
Dear Bart,
Thank you very much! I think it makes sense and the analysis seems pretty straightforward. I am sure to have seen this kind of comparisons in several papers but I cannot remember which ones. You are the expert indeed!
PS: I hope everything is fine with you. Best regards from Spain!
Dear Ruturaj,
Thank you for your answer. However, I don't think that repeated measures ANOVA is the right test. I have two independent groups and two independent measurements (so perhpas between-subjects designs are preferrable over within-subjects). Still, I think that if I just put together the conditions and run simple univariate ANOVAs I would surely incur in some kind of error. That's why Bart told me to take estimated marginal means.
I guess...
- Badges
- Science topic
- Similar topics
- Econ
- Business Economics