Hi everyone,
I am doing a quasi-experimental research to evaluate the effectiveness of a particular teaching method as a treatment in teaching vocabulary. In the treatment, I have 4 independent variables that can improve learners' vocabulary acquisition. I am planning to conduct overall pre-test and post-test as well as 4 individual pre and post tests for the different variables. Please help to explain the type of my quasi-experimental.
When you say "quasi-experimental" I assume that it means you are not using control groups.
It sounds like what you have in mind is Within Subjects or Within Groups design. See if you can find one of Creswell's books on research design.
You should have rendered more details regarding your sample and the how of conducting pretests and posttests. Assuming you have one group and a series of pretests and posttests along with various IVs, it could be interrupted time-series quasi-experimental design with multivariate analysis. You may check out Reichardt's (2019) guide.
Reichardt, C. S. (2019). Quasi-experimentation: A guide to design and analysis. Guilford Publications. https://books.google.com/books/about/Quasi_Experimentation.html?id=_z5ovgEACAAJ
Good luck,
Dear Michael W. Marek , I have two groups (experimental and control).
Dear Mohialdeen Alotumi , my participants were divided into control and experimental groups.
Zachary Farouk Chai If you have control groups, why do you call it QUASI-experimental? Why is it not regular experimental methodology?
Hi Michael Marek,
Can you explain your assumption of having no control groups in quasi-experimental designs?
True and quasi-experimental designs are majorly different in the way of selecting samples. That is to say, in quasi-experimental designs, the selection of the sample of the study is not random as it is in true experimental designs.
Thanks,
It seems that you are just using a before-after estimator. The assumption here is that the counterfactual of the post-treatment state is just the pre-treatment state.
Ali Alzubi If there are both control and experimental groups, why is it a quasi-experimental and not experimental design?
It is a quasi.... coz the selection of the sample of the study was not random.
In my opinion, If you do a quasi experiment, you do not have neither control nor experimental groups. You just conduct the research in one class, initiate the research by doing pre-test before treatment, giving treatment, and then finally doing post-test to evaluate the effectiveness of the treatment.
Zachary Farouk Chai It would be helpful if you could clarify what the four treatments are and why you need four tests. (I imagine that you have something like four sets of words and that you will teach each set in a different way, but I obviously can't be sure.) It would also be helpful if you could say what kind of statistical procedure you were think of using. (My guess is a MANOVA or Linear Mixed-Effects Model, but ...)
Martin Willis Thank your for your response. I have four different treatments (Word Recognition, Action, Picture and Contextualisation) to teach vocabulary. Hence, I need to evaluate the effects of these four treatments using different tests with different sets of vocabulary. I use Sample-T tests for the statistical procedure.
Zachary Farouk Chai First of all, I wouldn't recommend multiple t-tests. Linear mixed-effects modelling would be much better.
I'm still a bit unclear about the relationship between the words you will teach and test, and the treatments. Will each treatment have its own set of words? I.e. Word list A for Treatment A, Word list B for Treatment B, etc.. Or something else?
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity