What is pretest post test design of the research of psychology and which statistical tool suitable for it ?
Answer:
Well, this is usually used in clinical research for instance, when one wants to test the effectiveness of particular kinds of treatments (e.g. CBT alone, CBT with drugs, another therapy, control group...). In this case, you have a pre-treatment measure of the participant's anxiety/depression/phobia (wot u r measuring!), then have the intervention, and then after intervention, measure their anxiey/depression/phobia --- thereby getting a post-test measure.
As far as statistics are concerned, it depends. It usually would be a mixed between-within subjects ANOVA if you are looking to see the difference between the groups (i.e. the different interventions) as well as within a group (i.e. pre-test measure and post-test measure).
I hope I have made sense...since it's stats it's sorta harder to explain in writing!
There are many books in the library but the one that I have on my mind is call "Experimental method of statistics" it is very simple and well written, it has methods used in the field for analysing farm crops, but can be applied to psychology readily.
It has Large sample and Small sample theory and N x N and N x M method of analysis, with full explanations and examples. You need to pick one of the example and all you have do is, to put your data in same manner and follow the same steps.
There are many ready made programs, and the cheapest is hand held statistical calculator.
You can buy various program for the PC and or Mc. Buy it and use it, some are very user friendly.
Some of the programs are on the Internet in the public Domain, so it is a free down load.
I had to get a help of a programmer to program the whole thing in FORTRAN, and run my data, for analysis, it was not in psychology but you can find it at U Mass, Amherst, MA library, MS thesis titled "Electro magnatic radiation for Cockroach control" MS thesis Feb 1967 by Minoo Sumariwalla.
******************************...
There are series of books which are Known as schems(spellin?) series, they are in guide form, go to a university book stores and ask the people who work their they will have list of all the book(guides), there may be one that meets your requirements(medicle or and psychological statistics guide). Buy it. All books have list of other books listed on back cover page.
There is a book call Mathematical statistic by Fisher and an another auther, it has clinical test discussions and set of statistical set ups in it. May come handy.
Let's say you're a parent, and you have a messy kid who won't clean his room. Your best friend tells you that giving your kid lots of compliments, instead of yelling at him, will make him more likely to clean his room. You want to see if this is true.
So you set up a pre-test, post-test experiment.
For your pre-test, you ask your son to clean his room. On a scale of 1 to 10, in which 10 is the most clean, let's say that once he's cleaned it, his room is only a 3. (Very messy!)
Then, for the next 2 weeks, you do the "treatment" - you compliment him instead of yelling at him, just like your friend said.
At the end of the 2 weeks, to see if it made any difference, you ask him to clean his room again. This is the post-test. This time, on the scale of 1 to 10 (in which 10 is most clean) his room looks more like a 6! So you conclude that the treatment worked.
In other words, a pre-test post-test design is when you use the same group of participants, and test them at two different times to see if they changed because of something you made them learn or experience.
As to what test is suitable, this is actually a complicated question and based on what kind of data you have, what kind of variables you have, and whether your data falls in a normal distribution or not...but for your average research methods course, the right answer is a T-Test. You could also use a 2-group ANOVA, or a repeated measures ANOVA.
Best of luck!
The answers post by the user, for information only, FunQA.com does not guarantee the right.
More Questions and Answers: