Test for significance after intervention

#1
Hi all, newbie here who has been out of statistics for a while. I am attempting to test for two questions for an intervention that was conducted that involved the measurement of a test score prior to the intervention, monthly throughout the intervention, and after the intervention:
1. Is the reduction in score between the pre test and post test significant between observation and intervention group?
2. For the group in the intervention, does the length of time that they participated in the intervention lead to a more optimal post test score? I measured test scores monthly for six months throughout the intervention.

I thought that I should use a t test from the difference in means, but I am not feeling confident.

Also, there is bias in my groups in that they were not randomly selected, but rather self selected on whether they wanted to participate. With that in mind, is it even appropriate to measure significance?

Thanks for any help you can offer.
 

Karabiner

TS Contributor
#2
So you have 2 groups (how many participants in each group?),
and you measured each of them monthly, for 6 months? And the
intervention lastet 6 months for each participant in the
intervention group?

What is the dependent variable, and which instrument did you use to
measure it?

Dis you have study drop-outs or otherwise missing observations?

Did you record additional participant charcteristics (which
might be helpful to reduce possible bias due to self-selection)?

With kind regards

Karabiner
 
#3
Thank you for your response. I am new to this system and could only figure out how to attach my data with a picture lol.
-Dependent variable: Test score. The instrument I used is not known to others so not probably helpful to share. It essentially places each person on a scale of 0-5 based on activity outcomes.
-Observations: While the desire was for participants to continue for the full time period, there was great variability in how long they actually participated, as shown in my table below. Each column represent a unique population. This was implemented at the start of the pandemic, so I had challenges in completing my project. Even so, I would like to understand, if possible, if my results can have some confidence. I'd like to be able to say--"the longer they participated, the more optimal their results."

There are plenty of other characteristics that I collected, mostly demographic information. I am noting those differences and did apply a chi square to the categorical results so there is understanding between the two groups.

Does this help to clarify? Thank you once again for taking the time to help me.

Test Score Means Aggregate.JPG
 

Karabiner

TS Contributor
#4
What is the time between start and end in the Observation Group, is it 6 months?
What is the time between start and end in the Baseline Touchpoint group?

From a theoretical point of view and/or from the knowledge in this field of study:
are there additional participant characteristics which could explain pre-post
differences and should be included in the analysis in order to reduce bias?

With kind regards

Karabiner
 
Last edited: