I've got this!
What are three types of validity we learned (excluding internal and external). Explain how they are defined.
Content
Construct
Predictive
*Explanations will vary, but should be very detailed
Christen and Grace have conducted a study examining the effects of a special vaccine on herpes. In their experiment on a sample of 200 participants, the results are spotty at best: it works for some people but there is no predictable trend when it is retested. What is the validity of this test and which one(s) apply to this concept?
There is no validity that can be determined because the results are not reliable. While reliability is not essential for validity, it is difficult to determine validity if the results are not reliable. At best, we could attack the experimental design’s validity, but even then that’s difficult without reliability to do so.
I have a sample of 400 individuals. What is the margin of error?
1 / (square root of 400) = .05 or 5%
What is a stratified random sampling? Give an example.
*Answers will vary, but should touch on strata and taking subgroups for comparison.
How are mean, median, and mode arranged in terms of least to greatest for left skewed vs. right skewed?
Right-skewed
Mode ≤ Median ≤ Mean
Left-Skewed
Mean ≤ Median ≤ Mode
What is the difference between a confounding variable vs. a lurking variable? Give an example of each.
A confounding variable is something we can control for on the spot, we might predict it. However, a lurking variable is not determined until later (with a follow-up experiment usually). We did NOT see it coming because it was lurking like a creeper in the shadows of our design.
*Examples will vary, but should be very detailed
What is necessary for us to determine if a test has content validity?
Subject Matter Experts or judges for the questions to determine if they represent "accurate" measures
What is multi-stage random sampling? Give an example.
*Answers will vary, but should include taking a larger sample and progressively getting smaller and smaller (more narrowed after each stage of focus) until we have our sample of interest.
If a distribution is skewed, what two concepts can we use to measure the distribution of scores?
Median and Inter-quartile Range
A normal distribution is said to be symmetric and _______, with the _____, ______, and _____ being equivalent to each other. Typically the mean is ________, and the ________ is set at 1 when standardized.
Unimodal; mean, median, mode; 0, standard deviation
Keith and Marina are members of the same lab conducting a study comparing the results of experiments on four universities with a population of 1,000, 2,000, 3,000, and 4,000 students at each university, respectively. Using a random sample of 25% of students from each school, we know that the margin of error of each school is _______
Largest for the 1,000, followed by 2,000, 3,000, and 4,000
Megan and Brett conduct a study on workplace leadership (e.g., managers) by using intro to psych students, who are given course credit for participation. What type of random sampling should they use?
They cannot use probability sampling (everyone has an equal or calculated chance of being selected) because they used non-probability; specifically, convenience sampling. The chances of all intro to psych students having workplace leadership experience is very small and would not be an appropriate sample for this study.