This particular post will focus on the test for related samples (same participants tested in different conditions): its assumptions, calculation and significance testing.
what do I do with this??.. Today I will look into the Non-Parametric tests. I will start with a quick recap on what Non-Parametric tests actually are and when it is appropriate to use them. I will also remind how to rank data to make it ordinal and how to deal with tied scores.
This particular post will focus on the test for related samples (same participants tested in different conditions): its assumptions, calculation and significance testing.
0 Comments
I have covered parametric test for within-subject design. In this post, I will talk about the parametric test for between-subject design: Student's t-test for unrelated samples. Instead of looking at the differences between the pairs of scores (as t-test for related samples does), it looks at the differences between overall means of the two samples. I will cover the assumptions for this test, explain the calculation process and significance testing, and finally talk about the Levene's test, which is only implemented in t-test for unrelated samples. In the previous post I mentioned that t-tests are used to compare two means in order to reveal whether they are significantly different. In order to transform the means into t-scores and perform the mean comparison, t-tests are conducted. There are two kinds of those: between-subjects and within-subjects, depending on the measure design of an experiment (I will elaborate on this and give you some examples further down). In this post, I will talk about within-subjects tests, their assumptions, calculation and significance testing. Before that though, I will quickly explain the difference between Parametric and Non-Parametric Statistics tests. In this post, I will introduce you to t-scores and explain how they are similar to z-scores (remember those? They tell us how likely it is for a particular score to occur). I will also talk about t-distribution and its assumptions and explain the role which t-scores play in Hypothesis Testing; actual calculations and t-tests, however, will be discussed in the next posts. |
AuthorI am a 21 years old Psychology undergraduate in Edinburgh University. The idea behind this site is to provide some help to fellow students, to make studying psychology (including stats...) as much fun as possible and motivate me not to skip those 9am lectures! Archive
February 2013
Categories
All
|