Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
72 Cards in this Set
- Front
- Back
operational definition
|
1) defines variables with such specificity and concreteness that they make scientific research possible
2)specifies the procedures that researcher uses to observe the variable; both the independent and the dependent variables require operational definition |
|
nominal measurement
|
assigning numbers to categories that have qualitative rather than quantitative differences
|
|
ordinal measurement
|
data that differentiate between groups based on someone or thing having more or less of whatever is being measured than some other person or thing
|
|
interval measurement
|
data that have equal appearing intervals; scores of 1 and 5 are known to be 4 units apart in interval data
|
|
ratio measures
|
does that meet interval criteria, but also have a meaningful absolute zero point
|
|
reliability
|
the extent to which measurement yields numbers (data) that are consistent, stable, and dependable
|
|
instrumental error
|
in measurement, error that occurs because the measuring instrument was poorly written
|
|
application error
|
in measurement, error attributable to lack of control over how the measure was distributed or completed
|
|
random error
|
not a measure of anything
|
|
test-retest reliability
|
a method designed to assess reliability over time with the same test
|
|
internal consistency
|
the amount of agreement a measure's items have with each other
|
|
validity
|
the extent that scales or questions measure what they are thought to measure
|
|
content validity
|
whether a measure captures the content or the meaning of the variable being measured; an objective source of validity
|
|
face validity
|
whether a measures appear, on the face of itself, to be valid. a source of validity.
|
|
criterion-related validity
|
a measure is valid to the extent that it enables the researcher to predict a score on some other measure or to predict a particular behavior of interest
|
|
construct validity
|
whether a measure's items actually reflect what is being measured as determined by experts
|
|
pilot test
|
a preliminary data-gathering effort for the purpose of examining the research procedures, including the measures used, in order to correct any problems before the full study is conducted
|
|
Thurstone scale
|
a scale with predefined values associated with each statement. An equal appearing interval scales.
|
|
Likert-type scale
|
an equal appearing scale that asks participants to react to the statement on a range of responses from favorable to unfavorable.
|
|
Guttman scale
|
a unidimensional scale that assumes 1) unidimensionality and 2) that people, when faced with a choice, will also choose items less intense than the one chosen.
|
|
Semantic differential scale
|
a general scaling technique for measuring the meaning that an "object" has for an individual
|
|
response set
|
a pattern the respondent might get into based on simply marketing down one side or the middle
|
|
multidemensional scaling
|
a measurement scale that draws and anaolgy between a physical space and distances and a similar space and distances in our minds.
|
|
thermometer scaling
|
asks respondants to indicate their degree of like or dislike, favorableness or unfavorableness, agreement or disagreement on a 0 to 100 degree scale
|
|
open-ended question
|
a measurement technique where no predetermined response options are provided; questions asked in a survey are similar to essay or short-answer questions, allowing the respondent to answer in any way he or she chooses
|
|
unobtrusive measures
|
a measurement technique whereby the research passively observes behavior without the studied person's knowledge
|
|
demand cues
|
the sometimes subtle, unconscious cues that the researcher may emit to let the participant figure out what the researcher want to find
|
|
Hawthorne effect
|
a finding that resulted when research participants changed their behaviors because they felt they were being part of a study-- were being observed.
|
|
physical traces
|
an unobtrusive measure that examines the weathering of sites for behavior and use
|
|
archives
|
documents, as in historical/rhetorical-critical research; a technique found in unobtrusive measurement
|
|
observation (O)
|
in experimental design s used to refer to an observation, that is, a measurement
|
|
descriptive statistics
|
the reduction and simplification of data to describe, results
|
|
inferential statistics
|
statistical analyses that test if the results observed for a sample are indicative of the population; the presentation of information that allows us to make judgements whether the research results observed in a sample generalize to the population from which the sample was drawn
|
|
nonparametric statistics
|
statistical analysis where numbers resulting from nominal and ordinal measures are described; measures that represent categories of people, events, or things
|
|
parametric statistics
|
statistical analysis where numbers resulting from interval and ratio measures are described and inferences from samples to the population made; statistical techniques employed on interval and ratio data; measures that represent a continuum of data that at minimum appear to be equally distant apart
|
|
univariate statistics
|
descriptions or inferences about nonparametric or parametric variables by themselves
|
|
bivariate statistics
|
descriptions or inferences about two nonparametric or parametric variables as they relate to one another
|
|
multivariate statistics
|
descriptions or inferences about three or more nonparametric or parametric variables as they interrelate
|
|
frequency distribution
|
the number of responses that fall in a category in a nonparametric analysis; can be analyzed as numbers, percentages, or proportions
|
|
percentages
|
a nonparametric statistic of the number of each group divided by the total in all groups
|
|
proportions
|
a nonparametric statistic of the number of each group divided by the total for a finite number
|
|
contingency table
|
a nonparametric analysis where two variables are displayed
|
|
median
|
a parametric statistic, a measure of central tendency, the 50th percentile or midpoint of all responses
|
|
mode
|
a parametric statistic, a measure of central tendency, the most frequently occurring response; a variable may be multi-modal-- have more than response that is responded to equally
|
|
mean
|
the most common central tendency. the average-- the sum of all scores, divided by the number of scores.
|
|
range
|
a parametric statistic, a measure of central tendency, the difference between high and low scores
|
|
standard deviation
|
a parametric statistic, a measure of central tendency that tells us how dispersed the data points are from the mean
|
|
variance
|
a parametric statistic, a measure of central tendency that is a more general measure of dispersion from which the standard deviation is calculated, the square of the standard deviation
|
|
standardized score (Z-Score)
|
a parametric statistic, a way of describing a variable by translating each individual score into a score reflecting standard deviation units.
|
|
Spearman rho rank
|
a nonparametric statistic that measures the correlation between two categorial variables
|
|
edit
|
edit
|
|
statistically significant
|
a decision made based on the probability that a research finding is caused by the independent variable and not through random chance; typically set by convention at a 95 percent (or p < .05) probability level
|
|
null hypotheses
|
the actual testing relationship that no difference exists between conditions; statistical tests seek to reject the null hypothesis in favor of the research hypothesis of difference.
|
|
alpha (p level)
|
the significance level of an inferential statistical test
|
|
two-tailed test of significance
|
a statistical test for a hypothesis that predicts that some differences will be found, but without specifying how the differences will occur; if a= .05 ("one tailed") then 2a= 0.25 ("two tailed")
|
|
one-tailed test of significance
|
a statistical test for a hypothesis that predicts that difference will be of a certain magnitude and in a particular direction
|
|
type 1 error
|
error that occurs when a researcher claims that relationships exist when in fact they do not; an inappropriate rejection of the null hypothesis
|
|
type 2 error
|
error that occurs when a researcher claims that a relationship does, an inappropriate acceptance of the null hypothesis
|
|
degrees of freedom (df)
|
a correction factor in inferential statistics that allows you to predict one outcome by holding another constant (n-1)
|
|
independence
|
a statistical assumption that any participant selected in the sample does not preclude another participant's selection
|
|
normailty
|
a statistical assumption that the peakedness of the data distribution is normally distributed (as in the normal curve) and is the same for data in the population.
|
|
equality of variance
|
a statistical assumption that sample and population's data are equally distributed
|
|
chi-square tests
|
inferential statistical tests for non-parametric data
|
|
observed frequency (O)
|
in the chi-square nonparametric test, the number of objects or people who are placed in a given category.
|
|
expected frequency (E)
|
in the chi-square nonparametric test, the theoretically expected number of objects or people who are expected to be placed in a given category.
|
|
within-group variance
|
that part of the variance that is associated with error in inferential statistics
|
|
t-test
|
an inferential parametric statistic that tests for differences between two means
|
|
ANOVA
|
analysis of variance, a parametric statistics that tests for differences among means for groups
|
|
a priori method
|
use of reasoning to answer a question before a factual observation
|
|
post hoc method
|
"after the fact"; usually refers to statistics used after the primary analysis is completed
|
|
experiment-wise error protection
|
a multiple comparison inferential statistic that allows us to be sure that no matter how many multiple, post, hoc comparisons we conduct, the change that any one of them will be statistically significant remains the same. we are protected against type 1 error
|
|
skewness
|
positive skew: mean is higher than the median
negative skew: mean is lower than the median |