# Research Methods II

The flashcards below were created by user flashman on FreezingBlue Flashcards.

1. Type I error
the rejection of a true null hypothesis
2. Type II error
the acceptance of a false null hypothesis
3. Null hypothesis
a statement of no relationship between variables; rejected when an observed statistic appears unlikely under the null hypothesis.
4. Sampling Distributions
A theoretical distribution that can be specified for any statistic that can be computed for samples from a population
5. One Tailed Test
a statistical test where extreme results leading to rejection of the null hypothesis can be located at either tail
6. Two Tailed Test
A statistical test where extreme results leading to rejection of null hypothesis will be located at both left and right tails
7. What can happen when you control for a third variable in a bivariate relationship?
• 1. Nothing
• 2. The correlation between X and Y reduces dramatically (Either there is spuriousness or Z is intervening)
• 3. The relationship between X and Y changes but does not disappear (then Z is a conditional variable)
8. Issues in Questionnaire Construction
• What do you want to know? What do you expect respondents to know?
• Question Content (Fact or Opinion, wording, provide middle alternatives)
• Question Type: Closed vs Open-ended
• Response Format for Closed-ended
• Question order
9. What to consider when evaluating a Poll (1)
• Topic - is it something people know about?
• Screen for non-attitudes – Did the organization give respondents middle
• alternatives?
• Question wording -Did the organization give you the question wordings?
• Question Order: Questions can be placed to lead respondents in certain ways
10. What to consider when evaluating a Poll (2)
• Sample definitions - Probability sample? From what population?
• Sub-sample sizes and error margins
• What happens if certain demographics are pulled
• out? (error margin increases)
• What's the spin?
• Less expensive
• May be the only source
• You can replicate/triangulate someone else's research with your own
• May only approximate the data you want
• May have hidden biases
• Sometimes access is a problem
• Higher response rate
• Can probe and clarify
• Quick turnaround
• High costs
• Possible interviewer bias
• No anonymity