Question

Instructions: Read each scenario and use the checkbox list to identify the most likely measurement

issue. Briefly explain your choice and propose a solution.2. TheTupelo Housing Program assists economically disadvantaged and homeless individuals to move into safe, healthy, and affordable housing. National service volunteers counsel individuals on their housing needs, help them apply for housing assistance and follow up to provide continued assistance and to verify an individual’s housing status up to 9 months after initial service. National service volunteers find that it is more difficult to track homeless men than women over the 9-month period to verify their housing status, so outcome data are missing for many homeless male clients. What is the most likely measurement problem? ☐Reliability ☐Validity ☐Bias Briefly explain your answer. Briefly, how might the program address this measurement problem?

Answer 1

Answer:

Explanation: The absence of outcome data from a substantial subgroup of the service population (homeless men) means that results will underrepresent individuals in this subgroup. Since this subgroup is also less likely to achieve the intended outcome, results are likely to be biased towards overestimating the effectiveness of the program.

The measurement problem is not reliability because there is no indication that the measurement process produces inconsistent results. If anything, results may tend to appear more consistent as the systematic (but unintentional) exclusion of a key subgroup from the dataset reduces variation in the results.

The measurement problem is not validity because there is no indication that the measurement process fails to accurately measure the intended outcome (participants move into safe, healthy, affordable housing) for those who the program is able to measure.

How the program might address this measurement problem: Any solutions to the measurement problem will involve finding ways to get outcome data from more homeless men so they are adequately represented in the dataset. Any reasonable sounding solution addressing the logistical problem of maintaining contact with homeless men during the 9-month follow-up period would be relevant to solving the measurement problem.