Chapter 9 - Survey Methods Flashcards
(19 cards)
Survey
Set of questions or statements given to measure someone’s attitudes, beliefs, values, etc.
Psychological Tests
Types of surveys that’s been used for years and are considered reliable and valid.
Open-Ended Questions
Allows the participant to write their own answer.
–> Great for a wide array of responses.
–> Difficult to score with so many options.
–> Good for pilot studies, where top answers become the choices in the final study’s question.
Closed Questions
Gives the participants a limited # of selections and ask them to choose one or more.
Likert Scales
Asks the participant to indicate their level of agreement with a statement.
Response Acquiescence
Problem with Likert scales where participants tend to agree with statements.
–> Can be mitigated by balancing favorable and unfavorable options.
Demographic Information
Characteristics of your participants (ex. age, gender, annual income, area of residence, etc.).
Double-Barreled Question
Two questions in one.
Example: “Is it wrong to curse and drink alcohol?”
Leading Questions
When the way you word a question has a BIG impact on how people will respond.
Self-Selection Bias
When those who reply are probably different from those who ignore the request (ex. personalities, strength of opinions about the topic).
Nonresponse Bias
When people who return surveys (via mail) differ from the people who get the survey but don’t respond (“non-responders”).
Correlation
Assesses the strength of a relationship between two variables with scores ranging from -1 to 1.
Correlation Coefficient
The number that indicates the size/strength (bigger = stronger) and direction of the relationship (-/+).
Scatterplots
Way to visualize the strength and direction of a relationship between two variables (and identify outliers).
–> The closer to a straight line, the stronger the relationship (more variability = more spread out points).
Regression
Procedure used to predict a score on one variable from a score on another variable.
–> Regression line/line of best fit is calculated based on the scatterplot data
Multiple Regression
Performs the same procedure as a regression using MULTIPLE variables to predict a single outcome.
Mediator
Type of third variable that explains the relationship between two variables.
Moderator
Type of third variable that’s a particular situation when a relationship occurs.
–> This in correlational data is equivalent to an interaction in experimental data.