Nitpick: I think it is good form to give P values for things you think ARE coincidences as well. So we can compare the “coincidental” and the “causal” P values or whatever. Example would be control before vs control after P value.
-> Control before vs control after p values are provided (not looking good, p value within group alone is insufficient, can capture learning, hence why I do a between group % change test)
-> Control before vs after for verbal IQ (significant—learning effect / motivation / shorter time between tests ?)
-> Intervention before vs after for verbal IQ (no significant—backs up hypothesis that this works for fluid IQ only, and the control effect is learning + the advantages on time & motivation)
Nitpick: I think it is good form to give P values for things you think ARE coincidences as well. So we can compare the “coincidental” and the “causal” P values or whatever. Example would be control before vs control after P value.
I agree, that’s why I did that :)
Within the article you can find examples of that:
-> Control before vs control after p values are provided (not looking good, p value within group alone is insufficient, can capture learning, hence why I do a between group % change test)
-> Control before vs after for verbal IQ (significant—learning effect / motivation / shorter time between tests ?)
-> Intervention before vs after for verbal IQ (no significant—backs up hypothesis that this works for fluid IQ only, and the control effect is learning + the advantages on time & motivation)
One day I will learn to Read To The End Before Commenting. You really have done your due diligence.