# late for Stats Ex

PSY 223 Choose Your Test

Use this document in Milestone Two to select a statistical procedure for your final project. Beginning in

Module Five, you are going to learn about these in more depth.

Remember: A sample is a population subset. Each data set for the final project includes one or two samples.

Numbers are already assigned to the elements of a sample; so when it says “sample,” you are looking for a

group of measurements.

1. Does the research question for your data set indicate (a) there is one sample and (b) the question

aims to find a relationship between two variables?

You want a correlation. A correlation determines the magnitude and direction of relationship

between two variables (e.g., amount of sleep 30 participants got and length of time they spent on

treadmill next day).

2. Does the research question for your data set indicate (a) there are two groups and (b) the question

aims to find out if the groups differ in some outcome or performance?

You want a two sample t-test. It reveals if the averages of the two groups are “different enough” for

us to conclude that the difference in the way the groups were treated led to the average difference

(e.g. one sample drank coffee, one sample drank decaffeinated coffee, and their average reading

speeds were compared to see if a difference was significant).

3. Does the research question for your data set indicate (a) there is one group and (b) the question

aims to find out if the group’s outcome or performance differed under two conditions?

You want a paired sample t-test. It shows if the average performances of the group under separate

conditions are “different enough” for us to conclude that the different conditions led to the

different performances (e.g. a sample of individuals learn a set of 10 words under influence of an

experimental memory drug; two days later, the same sample learns a set of words without any drug,

and their performances are compared).

USE DATA FROM SCENARIO 5 FRIDAY 13TH