PU565 Unit 4 Sampling Strategy Plan Example

Maximum Variation, Not Representativeness: A Unit 4 Sampling Strategy Plan Selecting Composite County Adults Over Sixty Across Three Dimensions, With the Case for Twenty-Four Interviews and Against the Three Alternatives

[Student Name]

Purdue University Global

PU565: Qualitative and Quantitative Methods

Unit 4 Sampling Strategy Plan

[Instructor Name]

[Date]

Culbreth County and its recruitment sites are composite; the sampling logic and the research cited are real.

What This Sample Is For

The interviews exist to find the range of ways people in this county account for not having a hearing aid. A sample that reproduced the county's demographics would answer a different question, how common each account is, which the later survey is designed for. Purposeful selection for variation is the appropriate strategy when the aim is coverage of a phenomenon rather than estimation of a parameter (Palinkas et al., 2015).

The Three Dimensions

1. Severity of self-reported difficulty: mild, moderate, severe, by a published screening score band rather than by the participant's adjective.

2. Social exposure: lives alone, lives with others, or has substantial caregiving responsibility for somebody else. This dimension is included because the account somebody gives of a device is partly an account of who will see them wearing it, and the stigma literature in this field reports exactly that audience effect (Wallhagen, 2010).

3. Previous contact with hearing care: never tested, tested and advised to get an aid, or tested and told to wait. The third category matters most and is the hardest to recruit, since those participants have a clinician's statement to reason against.

The Matrix and the Target

Three severity levels by three social categories gives nine cells, with the contact dimension layered across them rather than crossed, since a fully crossed design would require 27 cells and recruitment the study cannot deliver. Target: at least two participants in each of the nine cells, 18 minimum, with six further interviews held in reserve for cells that turn out to be heterogeneous. Total 24.

Why Twenty-Four

Three reasons, and the honest one comes first. Twenty-four is what the study can recruit and interview in nine months with one interviewer. Second, the empirical work on how quickly new codes stop appearing in homogeneous samples has found most codes present within the first dozen interviews, with later ones adding detail rather than new categories, which makes 18 to 24 defensible for a sample deliberately spread across cells (Guest et al., 2006). Third, the reserve of six is the part that protects the design: if the lives-alone cells produce accounts that diverge sharply, the reserve goes there rather than being distributed evenly. The number is not a target for saturation. It is a budget, with a stated rule for where the last six go.

Recruitment Routes, and the Selection Each One Imposes

1. Two county senior centers. Reaches people who leave the house and join things, which is a bias toward social exposure and against the most isolated.

2. A church bulletin in each of three congregations. Reaches a different social network, including people who do not attend senior centers.

3. A notice at the two pharmacies that stock the devices. Reaches people who have at least looked, which is the group the study most needs and the one least likely to volunteer.

4. Referral from participants, limited to one referral each, to avoid recruiting a single network's shared account.

No recruitment through clinicians, which would select for people already engaged with hearing care and would make the third contact category dominate.

The Three Alternatives, and Why Each Was Rejected

1. Convenience sampling at one senior center. Cheapest and fastest. Rejected because the accounts of people who attend a senior center daily are the accounts the county already hears, and the study would confirm them.

2. Snowball sampling from three initial participants. Good for hard-to-reach groups and wrong here, because non-use is not a hidden identity and a chain would deliver one social circle's shared reasoning as though it were a range.

3. Quota sampling to county demographics. Attractive to funders and misleading, since a demographically balanced sample of 24 cannot support any claim about proportions and would invite exactly that claim.

Stopping Rules

Recruitment in a cell closes when two interviews are complete and a third adds no account the first two did not contain, judged by the interviewer and recorded in a one-line note. Recruitment overall closes at 24 or at the end of month nine, whichever comes first, and the plan reports which it was, because a study that stops at its budget should say so rather than claiming it stopped at saturation.

What This Sample Cannot Support

It cannot support any statement about how common a reason is in the county, which is the sentence most likely to be written by accident in the discussion section. It cannot support comparison between the cells, since two participants per cell is coverage rather than contrast. And it will underrepresent the most isolated residents, whose accounts may be the most important, because every route above requires leaving the house or reading a notice. The survey phase addresses the first limit and not the third, and the plan recommends that the final report say so plainly rather than treating the interviews as if they stood for the county (Palinkas et al., 2015).

References

Guest, G., Bunce, A., & Johnson, L. (2006). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1), 59-82. https://doi.org/10.1177/1525822X05279903

Palinkas, L. A., Horwitz, S. M., Green, C. A., Wisdom, J. P., Duan, N., & Hoagwood, K. (2015). Purposeful sampling for qualitative data collection and analysis in mixed method implementation research. Administration and Policy in Mental Health and Mental Health Services Research, 42(5), 533-544. https://doi.org/10.1007/s10488-013-0528-y

Wallhagen, M. I. (2010). The stigma of hearing loss. The Gerontologist, 50(1), 66-75. https://doi.org/10.1093/geront/gnp107