pearson stats questions Qustions &
Answers Correct!!!!
-2 log likelihood (ratio) test - ANSWERUsed in logistic regression, it is a form of chi-
square test which compares the goodness of fit of two models where one model is a
part of (i.e. nested or a subset of) the other model. The chi-square is the difference in
the -2 log likelihood values for the two models
A priori test - ANSWERA test of the difference between two groups of scores when this
comparison has been planned ignorant of the actual data. This contrasts with a post hoc
test which is carried out after the data have been collected and which has no particularly
strong expectations about the outcome.
Adjusted mean - ANSWERA mean score when the influence of one or more covariates
have been removed especially in analysis of covariance.
Alpha level - ANSWERThe level of risk that the researcher is prepared to mistakenly
accept the hypothesis on the basis of the available data. Typically this is set at a
maximum of 5% or .05 and is, of course, otherwise referred to as the level of
significance.
Analysis of covariance (ANCOVA) - ANSWERA variant of the analysis of variance
(ANOVA) in which scores on the dependent variable are adjusted to take into account
(control) a covariate(s). For example, differences between conditions of an experiment
at pre-test can be controlled for.
Analysis of variance (ANOVA) - ANSWERAn extensive group of tests of significance
which compare means on a dependent variable. There may be one or more
independent (grouping) variables or factors. ANOVA is essential in the analysis of most
laboratory experiments.
Association - ANSWERA relationship between two variables.
Bar chart - ANSWERA picture in which frequencies are represented by the height of a
set of bars. It should be the areas of a set of bars but SPSS ignores this and settles for
height.
Bartlett 's test of sphericity - ANSWERA test used in MANOVA of whether the
correlations between the variables differ significantly from zero.
Beta level - ANSWERThe risk that we are prepared to accept of rejecting the null
hypothesis when it is in fact true.
,Beta weight - ANSWERThe standardised regression weight in multiple regression. It
corresponds to the correlation coefficient in simple regression.
Between-groups design - ANSWERBasically a design where different participants are
allocated to different groups or conditions.
Between-subjects design - ANSWERsee Between-groups design.
Bimodal - ANSWERA frequency distribution with two modes.
Bivariate - ANSWERInvolving two variables as opposed to univariate which involves just
one variable.
Bivariate correlation - ANSWERA correlation between two variables.
Block - ANSWERA subset of variables which will be analysed together in a sequence of
blocks.
Bonferroni adjustment - ANSWERA method of adjusting significance levels for the fact
that many statistical analyses have been carried out on the data.
Boxplot - ANSWERA diagram indicating the distribution of scores on a variable. It gives
the median in a box, the left and right hand sides of which are the lower and upper
values of the interquartile range. Lines at each side of the box identify the largest and
smallest scores.
Box's M - ANSWERA statistical test which partly establishes whether the data meet the
requirements for a MANOVA analysis. It examines the extent to which the covariances
of the dependent variables are similar for each of the groups in the analysis. Ideally,
then, Box's M should not be significant. The test is used in MANOVA though its
interpretation is complex.
Case - ANSWERThe basic unit of analysis on which data are collected such as
individuals or organisations.
Categorical variable - ANSWERA nominal or category variable.
Category variable - ANSWERA variable which consists of categories rather than
numerical scores. The categories have no particular quantitative order. However,
usually on SPSS they will be coded as numbers.
Cell - ANSWERThe intersection of one category of a variable with another category of
one or more other variables. So if a variable has categories A, B and C and the other
variable has categories X, Y and Z, then the cells are Awith X, A with Y, A with Z, B with
, X, B with Y, etc. It is a term frequently used in ANOVA as well as with chi-square tables
(i.e. crosstabulation and contingency tables).
Chart - ANSWERA graphical or pictorial representation of the characteristics of one's
data.
Chart Editor window - ANSWERIn SPSS it is a Window which can be opened up to
refine a chart.
Chi-square distribution - ANSWERA set of theoretical probability distributions which
vary according to the degrees of freedom and which are used to determine the
statistical significance of a chi-square test.
Chi-square test, Pearson's - ANSWERA test of goodness-of-fit or association for
frequency data. It compares the observed data with the estimated (or actual) population
distribution (this is usually based on combining two or more samples).
Cluster analysis - ANSWERA variety of techniques which identify the patterns of
variables or cases which tend to be similar to each other. No cluster analysis techniques
are dealt with in this book as they are uncommon in psychology. Often factor analysis,
which is in this book, does a similar job.
Cochran's Q test - ANSWERA test of whether the frequencies of a dichotomous variable
differ significantly for more than two related samples or groups.
Coefficient of determination - ANSWERThe square of Pearson 's correlation coefficient.
So a correlation of 0.4 has a coefficient of determination of 0.16. It is useful especially
since it gives a numerically more accurate representation of the relative importance of
different correlation coefficients than the correlation coefficients themselves do.
Common variance - ANSWERThe variance that two or more variables share.
Communality - ANSWERThe variance that a particular variable in an analysis shares
with other variables. It is distinct from error variance and specific variance (which is
confined to a particular variable). It mainly appears in factor analysis.
Component matrix - ANSWERA table showing the correlations between components
and variables in factor analysis.
Compute - ANSWERIn SPSS, this procedure allows the researcher to derive new
variables from the original variables. For example, it would be possible to sum the
scores for each participant on several variables.
Condition - ANSWEROne of the groups in ANOVA or the t-test.
Answers Correct!!!!
-2 log likelihood (ratio) test - ANSWERUsed in logistic regression, it is a form of chi-
square test which compares the goodness of fit of two models where one model is a
part of (i.e. nested or a subset of) the other model. The chi-square is the difference in
the -2 log likelihood values for the two models
A priori test - ANSWERA test of the difference between two groups of scores when this
comparison has been planned ignorant of the actual data. This contrasts with a post hoc
test which is carried out after the data have been collected and which has no particularly
strong expectations about the outcome.
Adjusted mean - ANSWERA mean score when the influence of one or more covariates
have been removed especially in analysis of covariance.
Alpha level - ANSWERThe level of risk that the researcher is prepared to mistakenly
accept the hypothesis on the basis of the available data. Typically this is set at a
maximum of 5% or .05 and is, of course, otherwise referred to as the level of
significance.
Analysis of covariance (ANCOVA) - ANSWERA variant of the analysis of variance
(ANOVA) in which scores on the dependent variable are adjusted to take into account
(control) a covariate(s). For example, differences between conditions of an experiment
at pre-test can be controlled for.
Analysis of variance (ANOVA) - ANSWERAn extensive group of tests of significance
which compare means on a dependent variable. There may be one or more
independent (grouping) variables or factors. ANOVA is essential in the analysis of most
laboratory experiments.
Association - ANSWERA relationship between two variables.
Bar chart - ANSWERA picture in which frequencies are represented by the height of a
set of bars. It should be the areas of a set of bars but SPSS ignores this and settles for
height.
Bartlett 's test of sphericity - ANSWERA test used in MANOVA of whether the
correlations between the variables differ significantly from zero.
Beta level - ANSWERThe risk that we are prepared to accept of rejecting the null
hypothesis when it is in fact true.
,Beta weight - ANSWERThe standardised regression weight in multiple regression. It
corresponds to the correlation coefficient in simple regression.
Between-groups design - ANSWERBasically a design where different participants are
allocated to different groups or conditions.
Between-subjects design - ANSWERsee Between-groups design.
Bimodal - ANSWERA frequency distribution with two modes.
Bivariate - ANSWERInvolving two variables as opposed to univariate which involves just
one variable.
Bivariate correlation - ANSWERA correlation between two variables.
Block - ANSWERA subset of variables which will be analysed together in a sequence of
blocks.
Bonferroni adjustment - ANSWERA method of adjusting significance levels for the fact
that many statistical analyses have been carried out on the data.
Boxplot - ANSWERA diagram indicating the distribution of scores on a variable. It gives
the median in a box, the left and right hand sides of which are the lower and upper
values of the interquartile range. Lines at each side of the box identify the largest and
smallest scores.
Box's M - ANSWERA statistical test which partly establishes whether the data meet the
requirements for a MANOVA analysis. It examines the extent to which the covariances
of the dependent variables are similar for each of the groups in the analysis. Ideally,
then, Box's M should not be significant. The test is used in MANOVA though its
interpretation is complex.
Case - ANSWERThe basic unit of analysis on which data are collected such as
individuals or organisations.
Categorical variable - ANSWERA nominal or category variable.
Category variable - ANSWERA variable which consists of categories rather than
numerical scores. The categories have no particular quantitative order. However,
usually on SPSS they will be coded as numbers.
Cell - ANSWERThe intersection of one category of a variable with another category of
one or more other variables. So if a variable has categories A, B and C and the other
variable has categories X, Y and Z, then the cells are Awith X, A with Y, A with Z, B with
, X, B with Y, etc. It is a term frequently used in ANOVA as well as with chi-square tables
(i.e. crosstabulation and contingency tables).
Chart - ANSWERA graphical or pictorial representation of the characteristics of one's
data.
Chart Editor window - ANSWERIn SPSS it is a Window which can be opened up to
refine a chart.
Chi-square distribution - ANSWERA set of theoretical probability distributions which
vary according to the degrees of freedom and which are used to determine the
statistical significance of a chi-square test.
Chi-square test, Pearson's - ANSWERA test of goodness-of-fit or association for
frequency data. It compares the observed data with the estimated (or actual) population
distribution (this is usually based on combining two or more samples).
Cluster analysis - ANSWERA variety of techniques which identify the patterns of
variables or cases which tend to be similar to each other. No cluster analysis techniques
are dealt with in this book as they are uncommon in psychology. Often factor analysis,
which is in this book, does a similar job.
Cochran's Q test - ANSWERA test of whether the frequencies of a dichotomous variable
differ significantly for more than two related samples or groups.
Coefficient of determination - ANSWERThe square of Pearson 's correlation coefficient.
So a correlation of 0.4 has a coefficient of determination of 0.16. It is useful especially
since it gives a numerically more accurate representation of the relative importance of
different correlation coefficients than the correlation coefficients themselves do.
Common variance - ANSWERThe variance that two or more variables share.
Communality - ANSWERThe variance that a particular variable in an analysis shares
with other variables. It is distinct from error variance and specific variance (which is
confined to a particular variable). It mainly appears in factor analysis.
Component matrix - ANSWERA table showing the correlations between components
and variables in factor analysis.
Compute - ANSWERIn SPSS, this procedure allows the researcher to derive new
variables from the original variables. For example, it would be possible to sum the
scores for each participant on several variables.
Condition - ANSWEROne of the groups in ANOVA or the t-test.
