Each student should respond to the posting of another group (i.e. a group different than yours) to critically evaluate the proposed sampling plan of the selected group. Make sure to highlight strengths as well as ways to improve.

Problem Statement – Workers are impacted as social work agencies are adapting a managerialism approach to management.

Research Question – How does managerialism impact workers at social work agencies that have adopted this management style?

Research Hypothesis – Workers who are employed by social work agencies that practice managerialism will have poorer mental health outcomes and higher rates of burnout than employees who work in social work agencies that do not practice managerialism.

Study Population – New York City social work agency employees who work in a variety of agency settings (both agencies that practice managerialism and those that do not)

Why? – The researchers will have access to this study population due to geographic proximity.

Sampling Plan – Non-probability sampling – purposive sampling and snowball sampling

Step 1:  each member of the group to select 20 participations from their social work agencies

Step 2: each participant to answer questionnaires on the impact of managerialism or non-managerialism, and how the management styles affect their work.

Step 3: each participate to answer surveys to rate their mental health while at work and at home.

Step 4: each participate to be interviewed and discuss which kind of management style would best work in the workplace for better productivity and mental health.

Step 5: Ask each participant to connect us with other workers who might be willing to be participants in the study.

Pros – Purposive sampling will allow us to select workers who have experience working in agencies that practice managerialism and those who work at agencies that do not practice managerialism. It would also allow us to specifically select workers who have been at their respective agencies for enough time to have a good sense of the management style and to feel the impact of the management style on their mental health. This will also be time effective. Snowball sampling will allow us to connect with more workers than we immediately have access to, as the initial sample of workers likely will be able to connect us with other workers at their agencies who may have similar experiences as they do.

Cons – Because these are non-probability sampling techniques and are not randomized, the sampling error cannot be determined. We also cannot determine if the experiences of the participants we will interview is representative of the general population, because the participants were not selected randomly. We may select bias participants to better answer our research question.

3 comments

Example of a student response:

This is a very interesting topic! I’m intrigued and wouldn’t mind giving my own coworkers the surveys!! My one concern, same for my own study, is the relationships and prior knowledge or participants completing the surveys the researchers will have. This makes me think there may be some bias in who is approached to take the surveys. For example, if a researcher knows a coworker is burnt out, maybe that would %100 be a person to participate because most likely they will prove the hypothesis to be true.