COM 400
Terms
undefined, object
copy deck
- equipment may wear out, information gatherers may become inconsistent
- instrumentation, or instrument decay
- subjects who achieve either a very high or very low score on a test tend to regress to the sample or population mean during future tests
- statistical regression
- TRUE OR FALSE: Testing extremes is a good idea.
- FALSE! BAD IDEA.
- occurs when subjects drop out of a test over time
- experimental mortality
- because many studies use two or more sample groups, they must be basically the same
- sample selection
- questions used to verify subjects responses; asks questions in different ways
- cross-validation
- bias can enter through mistakes in observation, data recording, mathemetical computations, and interpretations
- experimenter bias
- subjects may answer in a way they think is normal
- evaluation apprehension
- to show cause and effect of variables the researcher must establish that one happened first
- causal time order
- when the study is done to different groups at different times, information may be leaked
- diffusion or imitation of treatments
- sometimes individuals in a control group work harder because they are deprived of something. to avoid this, contrasting groups are often used
- compensation
- control group subjects may lose interest because they are not in the experimental group
- demoralization
- one a basic research idea has been chosen, the next step is to ensure the topic has _____
- merit
- one of eight question to determine topic merit: is the topic too _____?
- broad
- one of eight questions to determine topic merit: can the problem really be _______?
- investigated
- one of eight question to determine topic merit: can the data be ______?
- analyzed
- one of eight questions used to determine topic merit: is the problem _______?
- significant
- one of eight questions used to determine topic merit: can the results of the study be ________?
- generalized
- one of eight questions used to determine topic merit: what _____ and ______ are involved in the analysis?
- costs, time
- one of eight questions used to determine topic merit: is the ______ _______ appropriate to the project?
- plan approach
- how well the results of a study can be generalized across ppoulations, settings and times
- external validity
- Cook and Campbell: 1. Use ______ samples
- random
- Cook and Campbell: Use _______ samples and ________ the study several times
- heterogeneous, repeat
- Cook and Campbell: Select a sample _________ of the group to which the results will be generalized
- representative
- Cook and Campbell: Conduct research over a _______ period of time
- long
- four types of replication
-
literal replication
operational replication
instrumental replication
constructive replication - involves the exact replication of the previous analysis
- literal replication
- attempts to replicate only the sampling and experimental procedures of a previous analysis
- operaional replication
- attempts to duplicate the dependent measures used in a previous study and to vary the experimental conditions of the original study
- instrumental replication
- tests the validy of methods used previously by deliberately NOT imitating the earlier study; tries to find if the fact in the previous study can be found with a completely different kind of study
- constructive replication
- full-service research supplier participates in:
-
the design of a study
supervises data collection
tabulates the data and analyzes the results - field supplies usually specialize in:
-
conducting phone interviews
mail intercepts
one on one interviews
recruting respondents
- describes how easy it is to find a qualified respondent for a study
- incidence
- percentage of contacts reached out of all contacts made
- gross incidence
- number of respondents who actually participate in a study
- net incidence
- equal interval scales, typically used to measure an attitude, interval level scale
- Thurstone Scales
- AKA scalogram analysis,items are ranged along a continuum that each successive item is more extreme
- Guttman Scale
- most commonly used scales in mass communication aka summated rating approach
- likert scales
- used to measure the meaning an item has to an individual; bi-polar adjectives are used: pleasant vs. unpleasant
- semantic differential scales
- test is ______ if it consistently gives the same results
- reliable
- if a test is ______, it is consistent at different points in time
- stable
- involves examining the items for consistency in a scale
- internal consistency
- scores on odd numbers should correlate with even numbers
- split-half reliability
- assesses the relative correlation between two tests measuring the same concept; difficult to develop parrallel ratings forms
- cross-test reliability
- when two or more coders judge the same content, used in content analysis
- intercoder reliability
- measuring device is examined to see if it is valid on the face of it
- face validity
- substituting other words or concepts for another word
- constitutive definition
- specifies procedures that allow one to experience or measure a concept
- operational definition
- statement of fact meant to explain, in concise terms, an action or set of actions that is generally accepted to be true and universal
- law
- kind of reliability that says the scores on odd numbers should correlate with even numbers
- split-half reliability
- this is a method of knowing
- tenacity
- probability samples are selected at ________
- random
- this method of research surveys the entire population
- census
- comparing two tests for reliability
- cross-test
- a kind of validity that looks at how well the results of a study can be generalized across populations, settings, and times
- external validity
- categories of data would represent what kind of data?
- nominal
- a kind of variable that is not discrete
- conintuous variable
- in regard to guidelines for surveys, you should make questions what?
- clear
- items are ranged along a continuum so that each successive item is more extreme
- Guttman Scale
- what level of data would rankings in a beauty pageant?
- ordinal
- a kind of research other than public
- private
- this level of data has a true zero
- ratio
- one early effect on mass communication research was the magic ______ theory
- BULLET