Soc 20 Midterm 2 SG

Sample

subset of the population; we generalize from samples to populations

Sampling

the process of selecting a smaller group (the sample) from a larger population to conduct research

Probability (Random) Sampling

Def: researcher draws a sample from a larger pool of cases or elements

Benefits: More representative of the population than are non-probability samples

Allow us to estimate the accuracy of our findings

Drawbacks: high costs and effort, potential for sampling bias, limited sample sizes

Non-probability Sampling

Def: selecting a sample from a population where the selection is not based on random chance, but rather on the researcher's judgment, convenience, or specific criteria

Benefits: diversity of representative samples makes detecting cause-effect difficult; we can gather more and better information on non-representative sample

Drawbacks: may not be representative of the population of interest, findings can usually only be considered “exploratory”. Possibility of bias, either conscious or unconscious - researcher exercises discretion in selecting subjects

Target Population

a subset of a population is selected for a study

Population Parameter

a characteristic of an entire group, such as the mean, range, or standard deviation, that is typically unknown and needs to be estimated from information gathered from a sample of that group

Sampling Frames

the list of the elements in the population – specific list that approximates the population

Unit/Element

elements/individuals from the population included in the sample

Random Selection

choosing participants from a population where each individual has an equal likelihood of being selected

Probability Samples

1. simple

2. systematic

3. stratified

4. cluster

Simple Random Sampling

 every element in the population has an equal chance of being selected

Stratified Sampling

dividing the population into distinct subgroups (strata) based on specific characteristics before sampling

Multistage cluster

you select entire clusters rather than individual members, making it particularly efficient for geographically dispersed populations implementing multiple levels of cluster selection.

Cluster Sampling

divides the population into naturally occurring groups or clusters — you select entire clusters rather than individual members, making it particularly efficient for geographically dispersed populations.

Non-probability Samples

1. convenience

2. purposive/judgemental

3. snowball

4. quota

5. deviant case

Convenience

based on availability rather than representativeness

• least defensible

• okay for piloting questions

Quota

Non-probability version of a stratified sample

• Research has knowledge of populations characteristics (race, gender, age, social class)

• Population is stratified along these dimensions

• Researcher uses a convenience sample to fill N= of each category/strata

Purposive/judgemental

Researcher uses his/her best judgment to select sample

• Need knowledge of the population of interest

• Not concerned w/ representativeness

• Used in exploratory or field research

• Selects unique cases that are especially informative

Snowball

chain-referral sample — sampling a network

• Each person or unit in the study is connected to another through a direct or indirect linkage

• Once an appropriate subject is identified, they are asked to recruit others from their network who meet the study requirements;

• Often used to populate sample of individuals from “hidden” populations.

Census

a study that includes data in EVERY member of a population, not just a sample

Poll

A very brief single-topic survey

Sample Bias

discrepancy between an assumed population's actual distribution of a specific trait and the degree to which it is present in a given sample — not representative of the entire population

Sampling Error

difference between the estimates from a sample and the true parameter that arise due to random chance

Sampling Distribution

the values for a variable in a subset of observations from a larger population

Oversampling

selecting respondents so that some groups make up a larger share of the survey sample than they do in the population

Primary Data Collection

social scientists design and carry out their own data collection

Secondary Data Source

already completed work of other researchers relating to your research

Respondent

The person who is interviewed

Key Informant

person who is usually quite central or popular in the research setting and who shares his or her knowledge with the researcher or a person with professional or special knowledge about the social setting.

Self-administered

A survey completed directly by respondents through the mail or online

Interviewer Administered

A survey completed directly by respondents and the interviewer in person

Response Categories

preset answers to questions on a survey

Attrition

loss of sample members over time, usually to death or dropout.

Open-ended questions

broad interview question to which subjects are allowed to respond in their own words rather than in preset ways

Close-ended question

A focused interview question to which subjects can respond only in preset ways.

Avoid double barreled questions

A question that asks about two or more ideas or concepts in a single question.

Emotionally Loaded Words

terms that carry strong associations with certain moral concepts, ideologies, and evoke strong emotions and imagery

Jargon/slang

specialized language or terms used by social groups to communicate more efficiently and clearly delineate who is a member and who is not

Double negatives

two negative terms are used in the same sentence — example: "I don't have no money"

Consider order effects

the order in which questions appear biases the responses

Social desirable responses

the tendency of individuals to underreport socially undesirable behaviors and attitudes, while overreporting socially desirable ones in survey responses

Screener questions

question that serves as a gateway to (or detour around) a follow-up question; also called a filter question

Skip pattern

A question or series of questions associated with a conditional response to a prior question

Format and Layout considerations

Modes of administration

Computer assisted

Mail questionnaire

Web questionnaire

Telephone questionnaire

Face-to-face interviews

Probes

Mutually exclusive and Exhaustive response categories

Composite Measures (Scales and Indexes; Likert scale)

Codebook — Response set — Building rapport

Refresh: cross-sectional and longitudinal surveys; repeated cross section, panel

Experimental group/condition (treatment group)

Control group/control condition

Comparison group

Matching

Pretest — Posttest

Double blind

Audit study

Cover story

Confederate

De-briefing

Selection bias

Experimental Designs

Strengths:

Weaknesses":

Causality

Experimental independent & dependent variables

Key components of a “true” experiment (classical design)

Pre-experimental design

Quasi-experimental design

Random assignment vs Random Sampling

Deviant case

Seeking cases that differ from the dominant population — goal is to collect “unusual”, different or peculiar cases & learn about “normal” but studying “abnormal”.