Korea Social Survey Analyst Grade 2 flashcards

20 cards. Read the term and the definition here. The deck at the top of this page is optional.

Operational definition

The specified observable procedure or indicator used to represent an abstract concept in a particular research study.

Sampling frame

The operational list or representation from which sampling units are selected, which may differ from the intended population.

Stratified sample

A probability design that divides the population into groups and selects units within each group according to the design.

Cluster sample

A design that selects groups of population units, with further sampling or complete observation inside selected groups.

Coverage error

Error arising because the sampling frame excludes eligible units, includes ineligible units or otherwise misrepresents the target population.

Construct validity

The extent to which evidence supports interpreting a measurement as an indicator of the intended theoretical construct.

Pilot study

A preliminary investigation used to assess procedures, wording and feasibility before the full research operation begins.

Codebook

Documentation describing variables, units, response values, missing-value codes and processing rules so another analyst can interpret the file.

Range check

A data-quality check that identifies values outside the permitted or plausible range for a specified variable.

Consistency check

A comparison of related fields to detect answers that cannot reasonably hold together under the survey definitions.

Unit nonresponse

Failure to obtain a usable response from a sampled unit, distinct from an unanswered item in a returned questionnaire.

Item nonresponse

A missing answer to a particular question within an otherwise participating unit’s survey response or data record.

Imputation

Assigning a replacement value to missing data under an explicit method whose assumptions and effects should be evaluated.

Interviewer effect

Variation in recorded responses associated with how interviewers contact, question, probe or record participants rather than the intended construct.

Standard error

The standard deviation of an estimator’s sampling distribution, describing uncertainty from repeated sampling under the assumed design.

Confidence level

The long-run proportion of intervals constructed by a specified valid procedure that would contain the fixed population parameter.

P-value

The probability under the null model of a test statistic at least as extreme as the observed statistic.

Type I error

Rejecting a null hypothesis that is true, with its probability controlled by the test’s stated significance framework.

Statistical power

The probability of rejecting a false null hypothesis for a particular alternative and specified study or testing design.

Residual

The observed outcome minus the value predicted by a fitted model for the same observation.