aim 1 recruitment and sampling Flashcards

1
Q

Define snowball sampling. Strengths/limitations, biases introduced from starting with site director and getting referrals from them? Biases from sites that refuse to participate. Different sampling option?

A

A snowball sampling method will be used in which existing study subjects will be asked to refer and recruit future study subjects. This is a non-probability sampling technique and therefore is subject to bias, but it is likely to enhance the richness of data collection as compared to a random sample.

Help facilitate recruitment- more likely to say yes to a colleague than me

Can intentionally say- not just people who are the most vocal or people who have had the most positive experiences, emphasize interest in diversity of opinions

Bias in starting w site director that they will refer me to the person who will say the most positive things (about the director, about the site)

Bias in sites that refuse to participate. More likely to hear data that is about positives and success. Acknowledge.

Maximal variation sampling- diverse chose who are expected to hold different perspectives

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Define extreme and instrumental cases

A

Instrumental: case picked for its potential to add more depth and nuance to what considering more of the majority of a site experience

Extreme: case picked for being an outlier or providing disconfirming evidence. Important to more holistically understand the program

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Why chose 4-6 cases? How select sites? What biases does that introduce?

A

Seemed like the minimum amount to address some of the strata I am interested in for sampling frame of cases: site tenure, site geography, site success

Balance of breadth and depth, interest in both what is the same and different among sites

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Inclusive of active and inactive sites confusing. Why? What are pros/cons? How will they be different?

A

A lot can be learned from failed sites. Big political and programmatic decision to close a site, understanding those challenges important for preventing them

Interesting to compare to sites that were not doing well but were able to “turn it around”

provide context for older more historic longitudinal data

Bias: recall longer, undercurrent to the conversation that they were a failed site, may change what they focus on and what they remember. Could be positive to be further removed from the program, may feel they are able to be more honest about challenges at the site or with the mayors office

How well did you know this?
1
Not at all
2
3
4
5
Perfectly