Week 7: Quantitative Designs Flashcards
What is a Likert scale, and what is it used for?
A set of categories or short descriptions that people can chose from; used for…
-Demographics (ex. gender identity: male, female, non-binary)
-Attitudes (ex. I enjoyed the service: very much, somewhat, not much, not at all)
What are the two types of survey design?
Cross-sectional: a one-time sampling from a population to learn about them or their experiences
Longitudinal: research that revisits its sample over time to learn about changes
What are some advantages of research surveys? (5)
-Flexibility
-Efficiency
-Generalizability
-Least intrusive and least time consuming for participants
-Structured; easy to prepare data for statistical analysis
What are some limitations of research surveys (2)
-Their generalizability has limits (due to low response rate, sampling error)
-Opportunity for inaccuracies (missing data, misunderstanding intent behind questions, testing effects, reactive effects skewing results)
How can a survey’s response rate be increased?
-Calling respondents in advance
-Making questionnaires attractive and easily readable
-Ensuring all responses are complete before a respondent leaves
-Providing a pre-stamped return envelope or anonymous drop box
-Providing an incentive for completion (gift card, lottery entry, etc.)
-Reminder messages
What are the two types of research survey design/delivery?
Indirect: mailed, online
Direct: telephone, in-person (group or individual)
What are some advantages (5) and disadvantages (2) of mailed and online surveys?
Advantages: self-administered, cost-saving, structured and easier to turn into statistics, one respondent at a time means less external influence, typically no consent form required
Disadvantages: low response rate decreases representativeness, no control over completion of items
What are some advantages (2) and disadvantages (4) of telephone surveys?
Advantages: interviewer can assist/clarify, higher response rate
Disadvantages: higher number of attempts to reach participants, not everyone has a listed phone number, interviewer bias, participants may feel coerced
What are some advantages (2) and disadvantages (3) of group administered surveys?
Advantages: multiple respondents at once reduces costs, higher response rate in-person
Disadvantages: impossible to protect anonymity, travel/parking/childcare issues, respondents may feel coerced
What are some advantages (4) and disadvantages (2) of in-person interviews?
Advantages: higher response rates, greater length and depth/complexity, respondents’ interpretations can be clarified, more personal than by phone
Disadvantages: requires careful training and supervision, potentially less valid/reliable findings
What is a secondary data survey?
A method of obtaining data from publicly available data archives, another researcher, or one’s own previous research on another topic
What are some advantages (2) and disadvantages (2) of secondary data collection?
Advantages: could save time and money in some circumstances, does not require ethical review
Disadvantages: researcher has no control over methods, sorting through missing data/unclear variable labels/other issues can be time consuming
What is meant by the notation O1 X O2?
O = observation, X = treatment
So O1 (pre-test) means we collected data,
X means we did a treatment,
and O2 (post-test) means we collected data after the treatment
What does internal validity mean?
The extent to which findings reflect a true phenomenon as opposed to some alternative explanation
What are some typical threats to internal validity? (5)
-Maturation/passage of time (conditions may have changed, making findings irrelevant)
-Testing effects (conditions of the test influencing behaviour)
-History
-Instrumentation (inaccurate, non-dependable, irrelevant measurement tools)
-Statistical regression to the mean (how extreme variables tend to shift closer to the mean over time)
What are the four essential elements of a randomized control trial?
- Experimental and control group
- Random assignment to groups
- Independent veriables manipulated by the researcher
- Pre AND post test
What are some examples of pre-experimental research design? (4)
-Survey
-One group, post-test only (one-shot case study)
-Longitudinal case study
-Post-test only with non-equivalent groups
What is required for a research design to be considered experimental? (2)
A pre-test and a post-test
What are some examples of quasi-experimental design in quantitative research? (2)
-Non-equivalent comparison groups
-Non-randomized pre-test and post-test comparison groups
What do R, X, O, and subscript numbers indicate in research design notation?
R = randomization
X = treatment/intervention
O = observation/measurement
Subscript numbers indicate repetition or measurement over time
What is simple random (probability) sampling?
-Research participants are chosen at random from a list of the whole target population
-All members of the target population have an equal chance of being selected
What is systematic random (probability) sampling?
-Research participants are chosen in a patterned, but still random, way (ex. choosing every 5th person on a list of the whole population)
-All members of the target population have an equal chance of being selected
What is stratified (random/probability) sampling?
-The target population is divided into subgroups (strata) based on certain characteristics
-An equal number of research participants are chosen from each strata at random
-Each member of a strata has an equal chance of being selected
What is cluster (probability) sampling?
-The target population is divided into subgroups (clusters); often based on geographical location (ex. university A, university B, university C), but individuals are not placed in homogenous groups based on personal characteristics
-Entire subgroups are then chosen at random to participate in the research
What is snowball (non-probability) sampling?
A recruitment technique in which research participants are asked to assist researchers in identifying other potential subjects
Often used when participants have an uncommon/rare quality that is essential to the research (ex. people with a certain genetic condition recruiting other research participants based on their support network)