SPSS Glossary
Terms
undefined, object
copy deck
- aggregate data
- data based on larger group from smaller
- alternative forms method
- measuring the same variable different ways with two equal forms
- analysis of variance
- comparing variance in the dependant variable in the three or more groups to the dependant variable
- antecedent variable
- independent variable that logically come before another independent variable
- association
- a change in one variable happens with the change in another variable
- authority
- accepted as fact based on the position of who tells you
- bar chart
- frequency distribution at the nominal and ordinal level
- benchmarks of acceptable risk
- accepted levels of risk for determining probability
- beta weight
- slope of a regression line after standardization
- bivariate analysis
- analysis of two variable simultaneously to determine if there is a relationship
- Boolean Connectors
- internet searches using and or not
- call numbers
- library numbers
- case study
- examining in detail on only one occasion
- causal chain
- variables linked together that explain a dependent variable
- causal hypothesis
- independent variable always changes another variable
- causation
- change in dependent variable caused by the independent variable.
- census
- gather data from every member of a population
- census data
- information collect by govt about entire population
- chi square
- cross tabulation data to determine significance
- classic experimental design
- random subjects with control group and experimental group
- cleaning data
- checking and correcting data
- closed ended question
- limited answers that they must choose from
- cluster sampling
- clumping population based on a criteria
- codebook
- equates answers to values for calculation
- coefficient of determination
- r2 variation of dependent variable explained by independent variable.
- coefficient of multiple determination
- r2 variation of dependent variable explained by all the independent variables in multivariate analysis.
- collapse a variable
- combine variables
- comparative value statement
- compare relative values of variables
- comparison group
- not randomly assigned control group
- compound hypothesis
- hypothesis with two or more variables
- concept
- organized characteristics of objects, phenomena and ideas
- conceptualization
- defining concepts so meaning are concise
- concurrent validity
- validity is demonstrated by measuring the predictor and criterion at the same time
- condensed table
- multivariate table with only one dependent variable being shown
- confidentiality
- information provided by subject will not be revealed
- confounding factors
- independent variables may affect dependent variables
- consistency bias
- people answering question answer questions in a way to sound logical and consistent
- construct validity
- extent in which a measure entirely covers a concept
- content analysis
- collect and analyze written and electronic media
- content validity
- measure covers the full meaning of a concept
- contingency questions
- filter question
- control group
- identical to experimental group but are not exposed to the experiment
- control variable
- constant variable used to determine relationship between two other variables
- convenience sampling
- haphazard samples
- criterion-related validity
- extent a measure is related to something it should be related to
- cross tabulation
- one variable is classified in terms of another
- cumulative percentage
- additive percentage of cases
- data archives
- collected data from previous research
- data file
- computer file that stores data
- data matrix
- template used to organized data
- debriefing
- explaining experiment to subjects after the experiment
- deductive reasoning
- generalizations are used to make specific statements
- dependent variable
- changes in a variable is effected by an independent variable.
- description
- information about a single variable or a single case
- descriptive question
- a question asking about reality rather than seeking an answer.
- descriptive statistics
- summary of values of a single variable measuring central tendency.
- deviation score
- difference between the value of a variable and the mean of the distribution, added together always equals zero.
- dichotomous variable
- variable with only two values.
- direction of a relationship
- how two variables are associated (positive or negative)
- directional hypothesis
- hypothesis that states how two variables are related.
- disproportionate stratified sample
- sample elements are disproportionately represent when compared to the population.
- ecological fallacy
- stereotypes used to make decision about individuals
- elaboration
- examining effects of control variables on the relationship of two variables.
- element
- something that a researcher collects information on.
- empirical political analysis
- a method of testing one ideas about politics by collecting/analyzing data.
- empirical question
- a question that can be answered by gathering information and analyzing that data.
- error component
- Unexplained variation of the dependent variable in a regression equation
- evaluative statement
- statement made based on personal experiences.
- event statistic
- data used to collectivities such as countries or cities.
- expected frequency
- expected frequency if variables are unrelated.
- experimental group
- group in an experiment exposed to the experiment
- experimental mortality
- loss of members for control and experiment groups over the length of the study.
- explanation
- explanatory question
- A question about why something exists or the cause of some behavior
- extensive approach
- getting a small amount of information from a large sample
- external validity
- results of an experiment can be generalized to other groups of situations
- F-Test
- statistical significance among the means from three or more analysis groups.
- face validity
- makes sense based on common definition
- factorial design
- design in which more than one independant variable is employed
- faith
- knowledge not based on proof
- field experiment
- real world setting
- filter question
- screening question
- frequency distribution
- value present in numbers and percentages
- frequency polygon
- line graph with interval/ratio on X axis and frequency/proportion on Y axis
- gamma
- a statistic that indicates the strength and relationship of two variables
- histogram
- bar chart where width measures interval and all bars touch which measures duration.
- history
- life events that effect the way people may answer a question.
- hypothesis
- statement of relationship that can be tested
- hypothetical-deductive system
- observation results in generalizations about the world
- independent variable
- the variable that is not effected by the dependent variable.
- index
- how you measure indicators to determine a concept
- indicator
- observable evidence
- inductive reasoning
- specific observation used to make a generalization
- informed consent
- telling the subject what you are going to do an them agreeing freely
- institutional review board
- reviews research to protect humans
- instrumentation
- developing measurement devices
- intensive approach
- lots of quests to a small group
- interactive effect
- effect of two or more variables on one dependent variable
- intercoder reliability
- reliability between codebooks
- internal validity
- how much changes in the indep variable changes the dependent variable.
- interquartile range
- range of value represented by middle 50%
- intersubjective agreement
- level of agreement between two people on how and event should be described
- interval level of measurement
- scale of measurement that not only rank orders but also shows equal interval between values
- intervening variable
- control variable placed between the independent and dependent variable.
- investment
- believing something because of personal stake in it being true.
- judgmental data
- measures that are based on the opinions of experts in categorizing units of analysis
- judgmental sample
- selected sample units believed to be representative of a population
- keywords and phrases
- lambda
- Stat that measures strength of association at the nominal level
- latent content
- deeper meaning of media messages grouped by content analysis
- levels of measurement
- amount of information and levels of precision
- liner regression analysis
- best fit line between multiple variables
- logic
- knowledge through mental thought
- macro level of analysis
- analysis of units for things that can be described geographically
- main effect
- direct effect of independent variable on dependent variables
- manifest content
- easily understood content of media messages
- marginals
- totals of columns and rows
- maturation
- change in subject to time
- mean
- a measure of the central tendency for variables
- measurement
- numbers are assigned units of analysis to determine worth
- measurement error
- difference be a concept and its operational definition
- measure of association
- strength and relationship between a variable
- measures of central tendency
- the most frequent value
- measure of dispersion
- distance from the central tendency
- median
- middle point of a distribution
- median group
- group that contains the median
- micro level of analysis
- focus on people vs countries
- modal category
- variable that contains the greatest number of cases
- mode
- the most frequently occurring variable
- multigroup design
- an experimental group is added for each independent variable above two.
- multigroup time series design
- design with multiple pre and post tests
- multiple correlation coefficient
- a measure of association between a dependent variable and tow or more independent variables
- multiple linear regression
- measuring the relationship between a dependent variable and two independent variables while holding one independent variable constant
- multistage cluster sampling
- groups are sampled in the early stages and individuals in the late stages.
- multivariate analysis
- looking at the relationships of two or more variables
- multivariate percentage table
- percentage table showing relationship between three of more variables
- natural experiment
- experiment where experiment and control group occur naturally in nature and are not chosen by the researcher.
- necessity
- changes in an independent variable are necessary to change the dependent variable
- negative relationship
- relationship is opposite
- nesting
- grouping multivariate analysis
- nominal definition
- definition using words assigned by the researcher
- nominal level of measurement
- analysis without ranking
- nonprobability sample
- nonrandom measurement error
- normal curve
- frequency curve with a bell shape
- normative question
- value laden question for evaluation
- null hypothesis
- no relationship between variables
- observed frequencies
- number of cases found in each cell of cross tabulation
- one-tailed test
- testing the directional hypothesis by comparing two means
- open ended question
- questions that allow people to write their own answers.
- operationalization
- defining concepts that allow them to be measured
- ordinal level of measurement
- classifies units of analysis into categories that have hierarchal order difference between the categories is unknown
- outlier
- plotted value outside the trend
- P-Sample
- the sample of persons analyzed in a study using the Q-technique
- parameter
- the boundaries of an analysis
- partial regression coefficient
- how much change in the dependent variable if you only change one independent variable and all the other independent variables remain constant
- percentage table
- display proportion as opposed to raw numbers
- percentile
- plagiarism
- population
- total set of units for analysis
- population element
- single member of a population
- positive relationship
- relationship between variable moving in the same direction
- posttest
- measurement of the dependent variable after being exposed to the independent variable
- prediction
- predictive validity
- prescriptive value statement
- a value statement characterized by words like should and ought
- pretest
- pretesting
- probability sample
- program evaluation
- proportionate reduction of error
- predicts relationship strength between two variables
- proportionate stratified sampling
- sample populations are selected based on known population elements
- Q-sample
- deliberate sample of statements taken from a group of communications
- Q-sort
- viewpoints selected from strong agreement to strong disagreement
- Q-technique
- data collection of human subjectivity
- qualitative research methods
- used to describe a situation in a natural surrounding
- quasi-experiment
- a sort experiment
- quota sampling
- sampling based on percentage appearing in nature
- random assignment
- random digit dialing
- random measurement error
- random error that can be expected in an experiment
- range
- difference between high and low variable
- ratio level of measurement
- has true zero point as well as all the lower levels of measurement
- recode a variable
- combining variables to simplify
- recording units
- reducing effect
- reducing the strength of the first two variables when a third one is added.
- refining effect
- the original relationship survives only when the third variable is controlled
- regression coefficient
- the slope of a regression line represented as b.
- regression line
- line that best fits a scatter plot
- reliability
- consistency of measurement
- replication
- third variable added does not change the first two.
- research design
- plan to show how hypothesis will be answered
- residual
- response set bias
- revealing effect
- reversal effect
- third variable changes the first two to opposites
- saliency bias
- early question shapes future response
- sample
- sample element
- sampling error
- sampling frame
- sampling interval
- sampling ratio
- scale
- scatter plot
- used to graphically show the relationship between dependent (Y axis) and independent X-axis) variables
- science
- secondary analysis
- self-administered questionnaires
- simple post design
- simple random sample
- slope
- shows change in indep and dependant variables
- snowball sampling
- small group recommends people with similar interests for sampling
- Solomon four group design
- adds two groups the classic experiment to limit bias from the pretest
- split-halves method
- splitting a group giving them the same test an comparing the results
- SPSS
- Statistical Package for the Social Sciences
- spurious relationship
- a relationship that appears casual but is the result of a third variable
- standard deviation
- measure distance from the mean
- standard error
- show sampling error
- standard error estimate
- square root of the average squared error
- standard score
- scores converted to standard deviation units
- statistic
- estimate of population parameter
- statistical significance
- likelihood that a relationship found exists in the general population
- stratified sampling
- strength of relationship
- how much constantly variables move together
- syntality variables
- Variables that describe the qualities of a region or group as a whole, such as American Cities that have city manager forms of government
- systematic sampling
- Sampling by selecting for inclusion every nth element on a list
- t-test
- A statistical procedure to determine the level of statistical significance between means from two analysis groups
- tautology
- a statement that is true by definition and cannot be confirmed or falsified
- temporal order
- Used to infer causality
- test-retest method
- double test at two different points in time to measure validity
- theoretical population
- the population about which the researcher want to develop generalizations
- theories of the middle range
- theories offering reasonable accurate explanations of a limited number of social behaviors
- theory
- something used to explain an occurrence
- time series graph
- two-tailed tests
- testing the null hypothesis by comparing the two means
- Type I error
- False positive
- unit of analysis
- false negative
- univariate statistics
- stats that describe a single variable for the purpose of description
- unobtrusive measures
- Measurement that does not involve direct interaction
- validity
- measure reflect concept to be measured
- variable
- unit of analysis that takes on more than one variable
- variable name
- variance
- measure of dispersion
- writing handbook
- Y-intercept
- where line crosses y axis
- z-score
- A standard score that indicates the number of standard deviation units that a score differs from the mean.