SOC 281 Chapter 4
Terms in this set (34)
-a mental image that summarize a set of similar observations, feelings, or ideas
-can mean different things to different people
-must be clearly stated by researchers
-is the process of specifying what is meant by a term
-must clearly state what you mean by your key concepts
-measurements need to be clear and consistent with the definitions you settled on
conceptualization in inductive research
an important part of the process used to make sense of related observations
conceptualization in deductive research
helps to translate portions of abstract theory into testable hypotheses involving specific variables
-number that has a fixed value in a given situation; characteristic or value that does not change
-ex. gender in study concerning all male fraternity
-components of a variable
-ex. variable is education --> indicator is years of schooling
social scientists follow this progression when moving from a vague understanding of what a term means to specific measurements in social research
-Nominal Definition - definition that is assigned to a term w/o any claim that the definition represents a "real" entity.
Operational Definition - definition that specifies precisely how a concept will be measured (operationalization)
Measurements - Careful, deliberate observations of the real world for the purpose of describing objects and events in terms of the attributes composing a variable
-the process of specifying the way that you will measure each concept
-social scientists typically operationalize concepts into one or more variables
ex. of concept - variable - indicator
-concept: binge drinking
-variable: frequency of heavy episodic drinking
-indicator: ask "how often w/in the past 2 weeks did you consume 5 or more drinks containing alcohol in a row?"
what to consider when choosing variables
consider purpose of research and time and resource limitations
2 criteria every variable must meet
mutually exclusive (every case can only fall into one category) and exhaustive (every case must be classified as having at least one attribute for the variable)
When deciding how to measure a concept, researchers must consider four things
-range of variation that interests them
-degree of precision needed
-whether the variable will be directly or indirectly observed
-type of variable or the level at which the variable will be measured
-research method for systematically analyzing and making inferences from text
-ex. observing ads in women's magazine over 25 years
-words or other features are coded to measure variables.
-after coding procedures are developed, reliability should be assessed by comparing different coders' results for the same variables.
-computer programs can aid in coding and improving coding reliability
closed-ended (fixed-choice) question
-survey question that provides pre formatted response choice for the respondent to circle or check
-when in doubt add "other" answer
survey question to which respondents reply in their own words, either by writing or talking
-composite measure based on summing, averaging, or otherwise combining the responses to multiple questions that are intended to measure the same concept
-Averages idiosyncratic answers to single questions.
-Must demonstrate patterned responses in pretesting.
-Must demonstrate consistency of responses.
-Many pre-existing indexes already exist (such as the CES-D); these have been pretested and promote comparability.
-Indexes are often presented in matrix format (a series of questions that concern a common theme are presented together with the same response choices).
-Indexes are usually calculated by summing or averaging
-composite measure based on combining the responses to multiple questions pertaining to a common concept after these questions are weighted differently, such that questions judged on some basis to be more important for the underlying concept contribute more to the composite score
-give more information
making observations as another way to operationalize variables
a.Good as a supplement to interviews and survey data.
b.Measurement of choice for research in natural settings.
i.Observers can't observe everything.
ii.Observations are filtered by observer.
iii.Presence of observer may change behavior.
-the use of multiple methods to study one research question
-it strengthens measurement.
-especially strengthens measurement when based on different methods of data collection.
-multiple measurement methods of the same variable strengthens validity.
-if measures diverge, measurement error is likely.
-if measures diverge, each measure may be measuring a different concept.
4 levels of measurement
nominal, ordinal, interval, and ratio
nominal level of measurement
-places cases into categories with no natural ordering
-must be exhaustive and mutually exclusive
-ex. state, ethnicity, religious affiliation
ordinal level of measurement
-allow us to determine the order of categories, but there is no quantifiable distance between the categories
-less than/more than
-must be exhaustive and mutually exclusive
-ex. coffee shop: small, medium, and large
interval level of measurement
-numbers represent fixed measurement units but have no absolute zero point, ranked by order
-addition and subtraction are possible.
-ratios are meaningless.
-there are few true interval level measures in the social sciences.
-indexes often create interval levels of measurement if the range of the index doesn't include a zero
ratio level of measurement
-numbers represent fixed measurement units where zero means absolutely no amount of whatever the variable indicates (an absolute zero point)
-multiplication and division are possible
-highest level of mathematical precision
-with this type of variable, we can rank cases, say how far apart cases are, and say that one value is X times as large as another value
-ex. numeric variable
measurement validity and its 3 types
-whether operations developed to measure variables actually do so
-face validation, criterion validation, and construct validation
-it just has to seem reasonable given what we know about the concept, or it has to be valid on it's face.
-ex: the frequency of attendance at religious services is some indication of a person's religiosity seems to make sense without a lot of explanation
-establishes that the results from one measure match those obtained with a more direct or already validated measure of the same phenomenon (the "criterion")
-ex. if Jason says he hasn't been drinking we can establish criterion validity by giving him a breathalyzer
type of validity established by showing that a measure is related to other measures as specified in a theory
two measurements of quality
-validity refers to the extent to which a measure actually measures what we think it does
-reliability refers to the extent to which a measure yields consistent responses on different occasions
-take the same measurement at two or more points in time
-ex. taking the same test multiple times and getting the same score would show reliability
-reliability achieved when responses to the same questions by 2 randomly selected halves of a sample are about the same
-if you get an A on the first half of the test you shouldn't get an F on the second half
-comparison of subjects' answers to slightly different versions of survey questions
-when wording of test questions are changed slightly your grade should stay roughly the same
-when using multiple items to measure a single concept, the degree to which the answers are associated with one another
-the stronger the association, the higher the reliability of an index
-ex. reliable if doing well on some questions matched up with doing well on others
-using two different observers to measure the same concept
-ex. whether your professor or TA grade your test your grade shouldn't be affected