NOTE: THE DEADLINE IS 7:25PM EST. THE POST WOULD NOT ALLOW ME TO POST THAT TIME ONLY 8PM WHICH WILL BE ALMOST TWO HOURS AND TOO LATE.
Scenario #1: Imagine you are a consultant for a large HMO. The CEO has asked you to design a program evaluation to determine if they are providing quality care to their patients. The next 15 questions (1-15) relate to Scenario #1.
- The CEO wants a qualitative study. Which of the following represents qualitative data?
Interview a random set of patients to assess their satisfaction with healthcare at the HMO.
Send a questionnaire to a random sample of patients with Likert-type rating scales.
Make objective observations using a checklist of patients’ interactions with the physicians.
Analyze the number of test procedures denied by the patients’ insurance providers.
The CEO wants to know the principal advantage of a qualitative study. Which statement is the more accurate?
A qualitative study will allow you to access a wider range of patient views.
A qualitative study will allow you to explore the patient views in greater depth.
A qualitative study will always provide more valid results compared to a quantitative study.
- What is the first step of the research process?Create a plan for researchSelect the research approachIdentifying the research problemObtain Human Subjects Approval
- If you decided to use an associational approach for the evaluation, which of the following conditions would apply?You would use random assignment of treatments or conditions to intact groups.You would use an attribute independent variable.
The outcome of an evaluation can only be assessed using a qualitative methodology.
The CEO wants the data to be representative of all the HMO patients. How would probability sampling achieve a representative sample of the patient population?
Probability sampling would survey every patient at the HMO.
Probability sampling would survey every patient at one of the HMO hospitals.
Probability sampling would survey a convenience sample of patients.
Probability sampling would survey a random sample of patients.
Which situation best describes a probability sample?
The researcher stands at the front entrance to a hospital and randomly interviews patients as they enter and leave the facility.
A researcher obtains a list of all of the patients who have been admitted to a hospital and selects ten percent of the patients to interview using a table of random numbers.
Snowballing is the best way to produce a representative sample.
Which of the following is the best way to ensure that the evaluation has high external validity (generalizability)?
Replicate the study at other medical facilities.
Use only reliable and valid measures.
Have all survey items reviewed by a panel of experts.
Perform a content analysis on all interview protocols.
The CEO is confused about the difference between external and internal validity. Which statement should you select to explain the distinction?
External validity deals with whether the results from a study generalize to other populations and situations, whereas internal validity deals with whether the conclusions from the study are logical and reasonable.
External validity deals with whether the measures used in a study are appropriate, whereas internal validity deals with whether the measures are stable and consistent.
Both internal and external validity are achieved when the sample of participants is large and they are randomly drawn from the population.
The CEO understands about different forms of bias and wants to minimize experimenter bias specifically. Which technique will best minimize experimenter bias?
Use a double-blind technique for interviewing.
Do not allow the researchers to analyze the raw data.
Survey a random sample of patients.
Control for all extraneous demographic variables in the study.
Two researchers are selecting participants for groups in a study. Which one is demonstrating the most experimenter bias?
Researcher #1 selects participants according to their height.
Researcher #2 selects participants according to the last four digits of their phone number.
Selection bias is a threat to the internal validity of a study.
Scenario #2: You are a manager in charge of the 360-degree assessment at a company that provides financial and accounting services to a wide variety of other organizations. The next 15 questions (16-30) related to Scenario #2.
You want your 360-degree assessment to have the highest level of measurement precision. Which of the following methods is most likely to achieve your objective?
Have each of the different people doing the assessment conduct interviews and then report their impressions.
Design a rating scale to be used and test it for reliability and validity.
Allow the people doing the assessing to devise their own rating system to ensure variety.
Have the raters focus on psychological traits of the employees (e.g., work attitude, personality, enthusiasm) rather than objective behaviors (e.g., production, work quality, time management).
If you were trying to explain the difference between a quantitative measure and a qualitative measure to one of your senior managers, which expression would be most accurate?
A quantitative measure examines behaviors in numerical degrees and a qualitative measure looks at behaviors in broad categories.
A quantitative measure is a more reliable measure of behavior and a qualitative measure is a more valid measure of behavior.
A rating scale (such as a Likert scale) used in a 360-degree assessment is an example of an ordinal measure.
- Assuming you want high validity to your assessment system, which of the following techniques is the best way to test for validity?Assess the inter-rater agreement.Perform a test-retest analysis.Check the inter-item consistency.Correlate the ratings with other measures of success.
Which of the following two techniques below is the best procedure for testing reliability?
Assess the inter-rater agreement.
Test if the ratings predict future job performance.
A 360-degree assessment instrument cannot have high validity if it is low in reliability.
- Possible problems with the a one group pretest-posttest design include
- group differencesPoor ecological validity.selection bias.
- If you wanted to compare men and women on a 360-degree assessment, you would use which approach?Descriptive.AssociationalComparativeQualitative
Sampling bias can reduce the external validity of a study.
You decide to evaluate your company’s 360-degree assessment using one of the four designs below. Which design will provide you with the most valid results?
Compare the performance of two divisions in the company, one using the 360-degree system and the other using the old system.
Compare the performance of all the employees before and after the introduction of the 360-degree system.
Measure the performance of two divisions in the company using the old assessment system. Have one division remain on the old system and the other division use the new 360-degree system and compare their performance one-year later
Randomly assign employees to two groups. One group is assessed using the old system; the other group is assessed using the new 360-degree system. Compare the performance of the two groups one-year later.
Which of the two designs is a true experimental design?
Solomon Four-Group Design.
Two Group, Pretest/Posttest Design.
The most crucial property of a true experimental design is the random assignment of participants to groups.
Let’s say you choose to compare two groups in the company. One group is allowed to submit a self-assessment and the other is not. What is the best way to deal with the possibility that these two groups do not start out as equivalent (e.g., they differ in the proportion of males and females, or there are salary differences, or they have different levels of educational achievement)?
Randomly assign employees to each group.
Select employees from two divisions in the company and matching them on demographic variables.
Use a convenience sample and give everyone a pretest.
Don’t worry about it because these differences will not invalidate your study.
Which of the following two groups is more likely to be equivalent?
Group A and Group B are selected by social security number—those with low numbers go in Group A, those with high numbers go in Group B.
Group A and Group B are selected by social security number—those ending with an even number go in Group A, those ending with an odd number go in Group B.
The problems with research designs that have nonequivalent groups can be solved if the two groups are given both a pretest and a posttest.
Scenario # 3: You are asked by the Dean of the College of Letters and Sciences at a major California university to assess the performance of virtual teams in their online classes. The next 15 questions (31-45) relate to Scenario #3.The Dean wants you to use a mixed methods approach. Which of the following best describes a mixed method?
A combination of individual and group interviews.
A combination of quantitative and qualitative measures.
A questionnaire with a mixture of Likert rating scales and summative rating scales.
A survey given to a mixed demographic.
As part of your research strategy, you propose to do focus groups with randomly selected virtual teams. Focus groups are classified as which type of research method?
Mixed method approaches to research are a bad idea because it is difficult to draw valid conclusions when researcher mix their methodologies.
Suppose you select as your dependent variable the number of posted comments made by each member of the online team. Which of the following would most likely be considered an independent variable?
The length of each posted comment.
The amount of critical thinking displayed in each comment.
Whether the team was large or small.
The degree to which team comments stayed on topic.
As part of your investigation, you wanted to know if the virtual teams worked better together when they were able to talk to each other using headsets as compared to simply using a discussion board. The comparison between using a headset and using a discussion board would be a(n) _____________ variable.
Another way to look at independent and dependent variables is to consider the independent variable as the cause and the dependent variable as the effect.
If your study of virtual online-teams was to test the relationship between class size and amount of team communication, which variable would you most likely want to control?
Type of class (e.g., science, math, philosophy, language).
Size of class (e.g., small, medium, large).
Number of posted comments.
Length of posted comments.
Which of the following is the best method for controlling extraneous (unwanted “nuisance”) variables?
- Randomly assign students to different teams
Use a large convenience sample.
Controlling some variables in a study is necessary because those variables may influence the dependent variable and limit what you can conclude about the effects of the independent variable.
You plan to study the effects of technology on the cohesiveness of the virtual teams. You’ve decided that you need a control (comparison, contrast) group for your study. Which group makes the most sense as a control group?
Teams that interact without the use of any technology.
Teams from social science classes.
Teams from humanities classes.
The Dean asks you why your study requires a control group. What is your best answer?
The control group establishes a baseline for comparison.
The control group increases reliability of the dependent variable.
A control group is typically a group that does not receive the experimental treatment in a study.
Your study will examine virtual teams over an extended period (at least several weeks). You understand that in any longitudinal study there is the possibility of subject attrition (e.g., members of the teams may drop the class and not be part of the team at the end of the study). Which technique, theoretically, best controls for attrition?
Permanently remove any team that loses a majority of its members during the course of the study.
Replace missing team members with new people who respond just like the missing members.
Whenever a team loses a member, randomly remove one member from every other team.
Do nothing. Simply report the losses then proceed as if no attrition had occurred.
The Dean wants to know why attrition is a problem. Which is the best answer?
Teams who lose members may be different from teams who do not (e.g., they may lose the weaker students and thus end up being “better” teams).
Teams who lose members contaminate measurement precision of the teams who do not lose members (e.g., smaller teams make it more difficult to assess larger teams).
One way to minimize the subject attrition problem is to maintain contact with the subjects and attend to their needs..
Scenario #4: You have designed a comprehensive survey of the sexual practices of readers of a major men’s magazine. The survey involves full disclosure, a signed a waiver, and then (1) a questionnaire that is administered by email, (2) focus groups of men recruited at several malls, and (3) individual interviews of men who volunteered to be interviewed in your private office The next 15 questions (46-60) relate to Scenario #4.
Which situation describes an instance of possible sample bias?
The questionnaire may have items that only apply to men.
The focus groups are allowed to vary in size.
The men recruited from shopping malls are not a representative sample of the readers of men’s magazines.
A private office is an environment that is too formal for investigating sexual practices.
Sample bias is a problem because
The researcher cannot be sure that the sample is representative of the population of interest.
Responses from the participant may not reflect their true beliefs or practices.
Sampling bias and selection bias are just different names for the same phenomenon.
- Which of the following is an example of an open-ended question?Tell me why you read this magazine?On a 1 to 5 scale, rate how much you enjoy this magazine.
- After you do your survey, you want to use a summated scale for a construct that has five items. This means that youSum the high and low questions.
Sum all of the five items.
- Sum the standard deviations within the five items.Sum the five items and divide by the standard deviation.