Common Myths
Define the term "scale" in counseling research
“A collection of items combined into a composite score and intended to reveal levels of theoretical variables not readily observable by direct means”
(p.220)
Name the two scale types
1. Criterion-referenced scale: Measures an individual's ability
2. Norm-referenced scale: Aims to differentiate individuals standing along dimensions of a given construct
(p.220)
According to the book, which has a higher survey response rate: paper or online?
Paper
(p.293)
What is the BASIC difference between a correlation and multiple regression?
Correlation= 2 variables
Multiple Regression= more than 2 variables
(p.295/297)
True or False: Cluster analysis is a more advanced statistical approach than Latent Class/Profile Analysis
False! Latent Class/Profile Analysis is more advances analysis
(p.308)
How long can it take to complete item construction?
Item construction can take up to six months to a few years to complete.
Item construction is not a linear process; researchers add, delete, and revise items multiple times.
Scale construction is a complex process which includes item generation, pilot testing, and, if necessary, translation and back-translation.
(p.221)
What does pilot testing ask the participant to do?
(2 major things)
1. Asking participants to respond to the items as a participant
2. Identify unclear or ambiguous elements about the items
(p.229)
What is stratified random sampling?
Dividing groups into smaller groups (Strata) based on attributes or characteristics, for example men and women. Then a sample from each group is extracted based on the proportional size of the group to the whole population.
(p. 292)
Questions involving ___ addresses "when" or "for whom" a variable most strongly predicts or causes an outcome variable
Moderators
"A moderator is nothing more than an interaction."
(p.300)
What are the 3 types of person centered research design?
1. Cluster Analysis
2. Latent class/profile analysis
3. Growth Mixture Modeling
(p. 306)
Is an extensive literature review important to item construction? Why or Why not?
The literature review is crucial to item construction because it can provide a solid theoretical/conceptual grounding of the construct and/or wealth of information from previous research findings that investigated similar constructs.
(p.221)
What two things should the researcher ask themselves about the sample during data collection?
1. Whether the sample is representative of the population of interest or a sample of convenience
2. Whether the sample is appropriate (typically at least 250-300 participants are needed for factor analyses)
(p.230)
What are the 4 major activities included in conducting survey research?
1. Developing research questions for a defined population
2. Developing and/or selecting the survey inventories
3. Selecting and developing a data collection method
4. Analyzing the data
(p.290)
A moderator is a variable that affects the ___ and/or ___ of the relationship between a predictor (IV) and a criterion (DV)
1. Direction
2. Strength
(p.300)
When indicators or variables used to make the classifications are continuous, latent ___ analysis is used
Profile
(p.308)
Does translating and back-translating items of a scale provide any estimates of the validity of the items/scales in another culture?
No, it only indicates that the items have been adequately translated into another language.
Thus, it is important to not only provide cross-cultural validation of a scale, but also seek to understand the culturally based meaning of that scale in different cultures.
(p.223)
What are the issues/disadvantages in having both positively and negatively worded items?
1. item polarity being reversed can confuse participants
2. items worded in opposite directions often perform poorly
(p.228)
What factors, based on design of surveys, limit the ability to have randomized sampling?
1. Accessibility to online surveys
2. Availability of time to be interviewed
3. Voluntary nature
(p.292)
What does the mediator establish?
The “how” or “why” one variable predicts or causes a criterion variable
(p.302)
When indicators are ___, latent class analysis is used.
Categorical
(p.308)
Name five of the seven common myths on scale construction?
1. Item construction can be done in a few weeks
2. Items can be easily constructed without an extensive literature review.
3. Use a convenience sample whenever possible
4. Factor analysis alone provides sufficient evidence of the scale's validity
5. A scale with strong psychometric properties developed in Western culture is universally valid
6. A literal translation ensures linguistic and cultural equivalence
7. Structural Elements of a scale, such as a likert rating scale are universal across cultures
What are the 9 steps of survey construction?
1. Conceptualizing and operationalizing the construct of interest
2. Conducting the literature review
3. Generating the items, indicators and response formats
4. Conducting content analysis and Pilot testing revising, and administering the items
5. Sampling and data collection
6. Translating and back-translating the scale, if necessary
7. Finalizing items and optimizing scale length
8. Testing the psychometric properties of the scale
9. Advanced item evaluation or refinement of the scale
How are descriptive and explanatory research different?
Descriptive research provides basic information about a variable or phenomenon (i.e the frequency of car accidents in a college town).
Explanatory research attempts to identify variables that explain the occurrence of a phenomenon (i.e beliefs about women, or beliefs about the acceptability of physical aggression).
(p.288)
What statistical analysis answers the question:
"How multiple predictors and dependent variables are associated in a single model using latent constructs estimated through multiple observed variables?"
Structural Equation Modeling
(p.295)
What are the 3 purposes of cluster analysis?
1. Exploration
2. Confirmation
3. Simplification
(p.306)