Question

6. For each of the scenarios below, identify the sampling blunder, speculate about the influence of the bias, and then make a recommendation for ridding the study of the biasing influence. a. A researcher wanted to know how people in the local community felt about the use of high-stakes testing in the public schools. The researcher spent the afternoon at Wal-Mart and randomly approached 100 shoppers to ask their opinion (they all agreed to cooperate). Random selection was accomplished with the use of a random number table (the numbers determined which shopper to target, such as the 16th to exit, then the 30th to exit, then the ninth to exit, etc.). b. A researcher wanted to know how students at a university felt about mandatory fees for all students to support a child care center for students with children. The researcher set up a table near the dormitory where many different types of students came and went. Those who stopped at the table and seemed friendly were asked to complete the questionnaire. c. To study differences in occupational aspirations between Catholic high school students and public high school students, a researcher randomly sampled (using school rosters and a random number table) 200 students from the largest Catholic high school and the largest public high school. d. To learn more about teachers' feelings about their personal safety while at school, a questionnaire was printed in a nationwide subscription journal of interest to many teachers. Teachers were asked to complete the questionnaire and mail it (postage paid) to the journal headquarters for tabulation. e. To study the factors that lead teachers in general to quit the profession, a group of teachers threatening to quit was extensively interviewed. The researcher obtained the group after placing an announcement about the study on the teachers' bulletin board at a large elementary school.

Answer 1

Scenario A -

Sampling the shoppers at a certain period of time (in this case, afternoon) can not be considered the representative sample for the population.

Speculation - The conclusion from the study can bend in oppose of high-stakes because of possibility that during afternoon mostly generation x or baby boomers visit the shopping center and they could tend to be more lenient towards students as compared to overall sample and thereby biasing the results toward opposing high-stakes testing.

Recommendation - Interviewing people at random time periods can mitigate the bias

For Scenario B-

Researcher wants to understand all of university's students feeling about mandatory fees but instead he is sampling students who are seemingly friendly.

Speculation- This can lead in biasing the conclusion towards feeling good about mandatory fees for all students to support a child care center for students with children. Because seemingly friendly students can be more likely to feel good about such a cause as compared to other students.

Recommendation- Setting up a table where all (not many) students can be seen and asking students at random about their feelings.

For Scenario C-

Researcher wants to understand differences in occupational aspirations between Catholic high school students and public high school students, so he should take a random sample from two groups separately which also mean he shouldn't choose students from one particular school.

Speculation - The occupational aspirations of largest catholic school students can be higher than the sample aspirations of catholic school whereas that of largest public school students can be lower than the sample aspirations of public school students.

This can result in comparison of the one end of the extreme of one population with the other end of the extreme of other population which instead should be comparison of the population's central tendency.

Recommendation- Sample random students from schools selected at random for both groups.

For Scenario D-

Teachers reading journals regularly can be the ones who are happier in their profession as compared to teacher's happiness overall.

Speculation- Sampling results only based on view of teachers who read journal regularly can be biased towards teachers feeling safe personally while at school.

Recommendation- Considering all mediums through which teachers can be contacted and then analyzing the results.

For Scenario E-

We intend to study the factors that lead teachers in general to quit the profession, but we have sampled teachers threatening to quit for interview.

Speculation- Teachers threatening to quit their job can bias the results towards certain factors such as less pay, non flexible timings etc.

Recommendation- Interviewing teachers who don't threaten to leave their job but politely want to leave their job along with teachers threatening to do so