T-test
Presented by Karen J. Cotter
The t-test defined
A t-Test (or Student's test) is an inferential statistic that is used to determine if there is statistical significance between the means and standard deviation between two groups or, did it happen by chance. Because it is not always possible or convenient to test an entire population, the t-test uses a smaller representative sample. The resulting t-score is a ratio of the difference between groups compared to the difference within groups (the larger the score, the bigger the difference between the groups. The larger the t-value the more likely the test is results are reproducible. This value is then compared to the p-value (the probability that the results occurred by chance) of the sample's t-test and a determination of statistical significant between the two tests can be made and is a reflection of the population, not just sampling error, or chance.
There are two types of t-tests. They are:
A independent (one-sample) t-test. This test examines the differences between the population means (µ) of two different (independent) groups.
A dependent (two-sample or paired-sample) t-test examines whether the means of a single population differ significantly before and after a treatment has occurred (pre and post treatment).
Calculating Dependent (paired-sample) t-test with SPSS
A paired-samples t-test was conducted to compare the pre and post blood pressure medication systolic readings. There was a significant difference in the scores for the blood pressure readings, post intervention (M=81.50, SD=9.19) and the pre-medication blood pressure readings had a (M=89.9, SD=14.02) conditions; t(9)=3.13, p=.012. Therefore the null hypothesis cannot be accepted and the results suggest that the medication can reduce systolic blood pressure readings.
Want more details about each study?
Independent t-test
(un-paired)Going to the movies can cost a lot of money. When choosing a movie that has a high Rotten Tomato score you know you probably get your money's worth of entertainment. Let's take a look at which studio produces the most movies with the highest Rotten Tomato scores. But does this represent a real difference between studios or is it just a coincidental streak of luck?
Dependent t-test
(paired, two-sample) Controlling high blood pressure is not always determined by eating right and exercising. Sometimes better living through chemistry, in the form of medication, can improve blood pressure. Let's take a look at how we would use a dependent t-test to decide if there is a statistically significant effect of a new blood pressure medication (pre and post testing) to reduce bp on a population.