By whom to measure social reality? Cheat Sheet (DRAFT) by faminconnue

- interviews or partic­ipation observ­ation
- Purpose: explor­ative and explan­atory
- Labour intensive: low numbers objects of study
- Often nonpro­bab­ility sampling
- In-depth insights

- Questi­onn­aires or/and surveys
- Purpose: descri­ptive and explan­atory
- With little effort high numbers objects of study
- Probab­ility sampling
- Genera­lizable

*Rather than studying the complete (research) popula­tion, social scientists mostly select (relat­ively few) people for study to discover things that apply to many more people who are not
studied, at least so can be claimed to a certain extent.*

in probab­ility sampling there are 4 ways in which objects of study can be selected:

1. Simple random: assign numbers to sampling frame and select numbers randomly (computer) -> each element has an equal chance of being selected
2. System­atic: every kth unit in a list is selected in the sample
3. Strati­fied: grouping of the units composing a population into homoge­neous groups before sampling.
4. Cluster: natural groups (clusters) are sampled initially, with the members of each selected group being subsampled afterwards

nonpro­bab­ility = chance to became part of a sample is unknown

- in general use:
→ Qualit­ative research
→ Popula­tions with no sampling frame

- Nonpro­bab­ility sampling, how are objects of study selected?

4 types:
1. Conven­ience: easy availa­bility of objects of study
2. Snowball: ask objects of study to suggest additional objects
3. Purposive: resear­cher’s judgment about which objects of study will be the most useful or repres­ent­ative
4. Quota: on the basis of pre-sp­ecified charac­ter­istics, so that the sample will have the same distri­bution of charac­ter­istics assumed in the population

statistic: the summary descri­ption of a variable in a sample, used to estimate a population parameter

Sampling error: the degree of error to be expected of a given sample instead of the population
→ decrease when sample numbers increase + when variations decrease

Confidence level: the estimated probab­ility that a population parameter lies within a given confidence interval

Confidence interval: the range of values within which a population parameter is estimated to lie

purpose of empirical research = repres­ent­ati­veness
→ and so, sample of objects of study from a population must contain the same variations that exist in the population
+ the quality of a sample of having the same distri­bution of charac­ter­istics as the population from which it was selected

method:
- sample of objects of study from a population must contain the same variations that existin the population
- the quality of a sample of having the same distri­bution of charac­ter­istics as the population from which it was selected