PSYC*2360 Chapter 7: Correlational Research and Survey Design Flashcards
What is a correlational study?
A research design that focuses on how variables relate to one another
T or F: Conducting a correlational study is not the same as calculating correlation.
True
Do correlational studies typically allow for the determination of cause-and-effect relationships?
No
What is a correlation?
A statistic that measures the linear relationship/ strength of an association between two variables
How is correlation typically denoted?
r
Which quantitative research strategy is typically employed for systematically collecting information from a group of individuals?
Surveys
T or F: Information obtained from surveys is typically generalized to a larger group of interest.
True
What is the difference between open-ended and closed-ended questions?
- Open-ended: A question that participants answer using their own words
- Closed-ended: A question that participants answer using a predetermined set of response options
Which measurement strategy assigns a number to represent the degree to which a person possesses or exhibits the target variable?
Scales
Why does a participant’s summated score on a scale provide a better measurement of the variable than a single question?
Because having multiple items tends to cancel out any random error
What is a summated rating/ Likert scale?
A scale where participants evaluate a series of statements using a predetermined set of response alternatives
How are items on a Likert scale phrased?
As statements to be judged rather than questions to be answered
What are double-barrelled questions?
Questions that prompt participants to provide a single response to an item asking two separate questions
What are response sets?
A response bias where the participant tends to give the same answer to most, if not all, items on a scale, regardless of the question being asked
What is an acquiescent response set?
A response bias where the participant tends to agree with most, if not all, items on a scale, regardless of the question being asked
What is the best way to avoid response sets?
Include reverse-coded items
Ideally, what proportion of items in a response set should be reverse coded?
About half
What is reverse-coding?
A scoring strategy where negative response alternatives are assigned higher numerical values and positive response alternatives are assigned lower numerical values
What is a forced choice scale?
A scale where a person must choose between only two response alternatives for each item
T or F: Forced response scales have high sensitivity in differentiating between options.
False. They have very low sensitivity.
What is an error of central tendency?
A response bias where participants tend to avoid using the extreme response alternatives on a scale
What is one way to improve precision and minimize errors of central tendency on scales?
Use more response alternatives
How many response alternatives are typically included in a Likert scale?
5-9
What is internal consistency reliability?
The degree to which the individual items on a scale are interrelated (measure the same underlying variable)
What is test-retest reliability?
The ability to produce similar measurements across time (the temporal stability of a measure)
In general, to properly evaluate test-retest reliability, how far apart should the tests be administered?
1-2 weeks
What is alternative-form/ equivalent-form reliability?
A form of reliability that evaluated how well a measure correlates with a similar, but different measure of the same variable
What is face validity?
The degree to which a scale appears, on the surface, to measure the intended variable
What are two potential consequences if the face validity of a scale is too high?
It may introduce demand characteristics and social desirability bias
What is content validity?
The degree to which items on a scale reflect the range of material that should be included
A unit test that doesn’t cover much of the material learned in class is poor in which type of validity?
Content validity
What is construct validity?
The degree to which the scale actually measures the desired construct
How is construct validity established?
By evaluating convergent and discriminant validity
What is the difference between convergent and discriminant validity?
- Convergent: The degree to which measures of a variable correspond to measures of other, theoretically related, variables
- Discriminant: The degree to which measures of a variable do not correlate to measures of unrelated variables
How can scales establish convergent and discriminant validity?
By including additional items to measure other related and unrelated variables
If a scale converges with measurements of similar variables and diverges from measurements of dissimilar variables, does it have high or low construct validity?
High construct validity
If a scale converges with measurements of dissimilar variables and diverges from measurements of similar variables, does it have high or low construct validity?
Low construct validity
What is criterion validity?
The degree to which a measurement relates to a particular outcome or behaviour
How is criterion validity established?
By evaluating the concurrent validity or predictive validity of the measurement
What is the difference between concurrent validity and predictive validity?
- Concurrent: The degree to which a measurement corresponds with an existing outcome/ behaviour
- Predictive: The degree to which a measurement corresponds with a particular outcome/ behaviour that occurs in the future
What is the difference between non-directional and directional hypotheses?
- Non-directional: Doesn’t make any specific prediction about how two variables are related
- Directional: Makes a specific prediction about the exact nature of a relationship between two variables
Why is it often best to include basic demographic questions towards the end of a survey?
If the participant decided to only complete half of the survey, they will likely have already answered the items that deal more directly with the research question
What is evaluation apprehension?
Anxiety or concern participants may experience about how their answers or behaviours may appear to the researcher
What are distractor items?
Items included in a scale to mislead participants as to the scale’s real purpose
By providing participants with the opportunity to sign up to receive emails about topics related to the study, which type of validity can be examined?
Criterion validity
What is area probability sampling?
A sampling strategy where researchers divide the population into subgroups based on geographic area, then randomly select participants from each
What is the response rate of a survey?
The proportion of the invited sample that actually complete the survey
Once responses from scales are collected, statistics are used to accomplish what four tasks?
- Evaluate internal consistency reliability of the scales
- Summarize and describe responses on each variable
- Measure the relationship between variables
- Decide if an association exists in the population as a whole and not just among participants
What is Cronbach’s alpha?
A statistic used to evaluate the internal consistency reliability of a scale
What does a Cronbach’s alpha of 0 mean?
There is no internal consistency among scale items
What does a Cronbach’s alpha of 1 mean?
All items measure the same thing and the scale is redundant
A scale with an acceptable internal consistency reliability has a Cronbach’s alpha of what value?
At least 0.70
What is variability?
The degree to which individual measurements of a variable differ from one another and the mean
What is standard deviation?
How much, on average, individual scores differ from the mean
Does a large standard deviation indicate a greater or smaller degree of variability between scores
Larger SD means greater variability
When both variables are continuous measurements, what type of correlation statistic is typically used?
Pearson r
In terms of associations, what does “linear relationship” refer to?
The extent to which the measurement on one variable can be used to predict the measurement on another
What is statistical hypothesis testing?
A procedure for evaluating the probability of obtaining certain results if, in reality, there is no association between variables
What does it mean for results to be statistically significant?
It is highly probable that the results obtained were not due to chance
