How to Compare Two Samples Using t Tests (11-1)

The most basic and most popular inferential statistic – especially for beginning statistics students – is the t test. The t test is wonderful because it allows us to compare two samples. This opens up a world of possibility for our research. Researchers often want to compare two (or more) sets of data such as comparing two different methods of teaching, or how two competing therapy techniques affect depression, or if changes in scores occur after a treatment. There are two ways to structure these tests: as separate samples or using same sample.

This video teaches the following concepts and techniques:
Comparing the mean of two samples for statistical significance

Link to a Google Drive folder with all of the files that I use in the videos including the Effect Size Calculator for t Tests and datasets. As I add new files, they will appear here, as well.

[StjepanBrbot]

There is an error in the slide: if population SD is known we would use z-test (not t-test), otherwise if population SD is unknown than we would use t-test. (in speech it has been said t-test for both).

[malvane]

The design you propose near the end of the video (laptops vs. pen/paper) wherein you split the year in half would only ever be done as a last resort option, after exhausting all possibilities of blinding the control from the intervention groups. Problem with giving them the intervention halfway is Regression to the Mean - they are cognitively not the same students at each point of the year.

[jamesleleji]

How do you conduct a difference in differences f

[fariaftab]

I am comparing immigrants and natives across the scores for neighbourhood embeddedness (1 low to 4 high). ttest shows that natives have 0.10 (mean for native 3.44 and migrant 3.34) points higher neighbourhood embeddedness. I want to communicate the substantivizes of the difference and so standardise the neighbourhood embeddedness. In the z-score form the mean for natives is 0.005 and for migrants is -0.116. Now I need to show that 0.10-unit differences in raw scores corresponds to X % of a standard deviation change in this variable?? How plz help

[rizwanalishinwari]

Hello sir,
I want to do a similar tests for Education Level and Conflicts impacts in Pakistan. What do you recommend which tests should I do, and which available variables should I take?