Glossary
Glossary
A
95% confidence intervals An estimate of scores that the true value lies within. In 95%
confidence intervals the probability of the true value falling within the limits is 95%.
analysis of covariance (ANCOVA) Assesses the impact of independent variables on a
dependent variable whilst controlling for the effects of extraneous variables that may have an
impact on the dependent variable.
analysis of variance (ANOVA) A statistical technique that is used in difference studies with
more than two groups. ANOVA calculates the F ratio to determine whether between group
variance is larger or smaller than the within group variance.
APA format Results written in accordance with the guidelines published by the American
Psychological Association.
assumption of equal variance Assumes that the variability in scores on continuous variables is
equal in all categories of discrete variables.
assumption of homoscedasticity Homoscedasticity is also known as homogeneity of variance.
Assumption that variance on variable x is equal at all values of variable y.
assumption of independence Assumes that scores on one case are not affected by scores from
any other cases.
assumption of multicollinearity Assumes that the independent variables are not highly
correlated with one another.
assumption of normality Assumes that data is normally distributed.
assumption of singularity Assumes that one independent variable is not a combination of scores
on other independent variables.
Back to the top
B
Beta values Show the magnitude and direction of the unique relationship between each
independent variable and the dependent variable in a multiple regression.
bivariate correlation Tests the relationship between two variables.
Back to the top
C
categorical data Also referred to as nominal data. Data belonging to this group can be sorted
into categories according to its value.
causality See cause and effect. Establishing causality seeks to identify a cause and effect
relationship.
cause and effect Establishing that changes in one variable caused a change in another.
central tendency A measure of the average score within a set of data, can be calculated in
several ways.
chi-square goodness-of-fit test Also referred to as the one-sample chi-square. A non-parametric
test used to compare the frequency distribution of cases on a single, categorical variable to
hypothesised values.
chi-square test for independence Also known as a chi-square contingency table analysis. A
non-parametric test used to establish whether there is a relationship between two categorical
variables.
confounding variable A variable that may impact upon your variables of interest.
continuous variable A variable that is measured on a scale rather than by placing cases into
categories or rank order.
correlation A relationship between two variables.
covariates A variable that is possibly predictive of an outcome. It has a relationship to the
variables of interest in your research design.
critical value The value that the test statistic must exceed in order for the result to be classed as
significant. The critical value varies depending upon the number of degrees of freedom and the
desired alpha level.
counterpart in the other group, i.e. someone with the same profile as them on key variables.
mean A measure of central tendency, the arithmetic average of a set of scores.
median A measure of central tendency, the middle number in a set of scores.
mode A measure of central tendency, the most frequently occurring number in a set of scores.
multivariate analysis of variance (MANOVA) A statistical procedure used to look at group
differences when there is more than one dependent variable in the research design.
Back to the top
N
nominal data See Categorical data.
non-parametric data Nominal, ordinal or interval/ratio level data that is not normally
distributed or does not possess homogeneity of variance is referred to as non-parametric.
non-parametric test Statistical analyses that do not require data to be interval level, be normally
distributed or have homogeneity of variance.
non-significant Used to suggest that the outcome of a statistical test shows an alpha level above
the accepted level.
normal distribution A distribution of scores which are mathematically defined and when
graphed appears symmetrical like a bell curve. Most scores are concentrated around the mean,
with very few cases at the extremes.
normality plots Graphical techniques for assessment of normality.
null hypothesis Hypothesises that there is no effect in the data, any patterns seen are due to
chance factors.
Back to the top
O
one-tailed hypothesis A hypothesis that predicts an effect in a particular direction.
ordinal data Scale where the relative position of points is given, but measurement points are not
evenly distributed. Position in a race is an example.
outliers Data points that vary from the general trend of the data. For example, singular very high
outputs.
Back to the top
R
random sampling Each case in the population has an equal chance of being selected.
range The difference between the highest and lowest score.
ranked data See Ordinal data.
ratio level data Scale where the measurement points are evenly distributed. Negative values are
not possible. Weight is an example.
residuals Difference between actual values and estimated values.
Back to the top
S
sample size Number of cases used in your study.
scatterplot A graph showing the distribution of scores on two variables for a data set. Each data
point shows the score for that case on both variables.
significant Used to suggest that the outcome of a statistical test shows an alpha level below the
accepted level. For example, in psychology the accepted alpha level is usually 0.05, giving a 5%
probability of the results being due to chance factors. Any test that returns the 0.05 or a lower
alpha level is termed a significant result.
skewed A data set is skewed if scores are clustered towards one or other end of the distribution.
Skews can be positive or negative.
Spearmans rho correlation analysis A non-parametric test used to establish whether there is a
relationship between two ranked or ordinal variables.
standard deviation A measure of the distribution of the scores around the mean.
standard error of the mean An estimate of the standard deviation of all possible population
means.
standard multiple regression A parametric test used to assess whether scores on a dependent
variable can be predicted from scores on multiple independent variables. Also allows assessment
Try the multiple choice questions below to test your knowledge of this Chapter.
Once you have completed the test, click on 'Submit Answers' to get your results.
Question 1.
Graphs which present the values on the horizontal axis and the number of times this
occurs on the vertical axis are known as what?
Line graph
Box-plot
Cluster graph
Scatter graph
Frequency distribution
None of these
End of Question 1
Question 2.
What would you use to compare the frequency distributions of more than one set of
data?
Frequency distribution
Frequency polygon
Scatter graph
Bar chart
Box-plots
Line graph
End of Question 2
Question 3.
What shape is a normal distribution?
Inverted (U)
Descending line
Bell shaped
Ascending line
Flat
Bi-modal
End of Question 3
Question 4.
What type of distribution is observed when most of the scores cluster around the
lower end of the scale?
A related distribution
A positively skewed distribution
A bi-modal distribution
A normal distribution
A flat distribution
A negatively skewed distribution
End of Question 4
Question 5.
What type of distribution is observed when most of the scores cluster around the
higher end of the scale?
A bi-modal distribution
A normal distribution
A flat distribution
A related distribution
A positively skewed distribution
A negatively skewed distribution
End of Question 5
Question 6.
What type of distribution has two peaks?
None of these
A positively skewed distribution
A bi-modal distribution
A normal distribution
A negatively skewed distribution
A flat distribution
End of Question 6
Question 7.
Which measure of central tendency is derived from the most common value?
Variance
Range
Distribution
Median
Mean
Mode
End of Question 7
Question 8.s
Which measure of central tendency is obtained using the middle score when all
scores are organized in numerical order?
Range
Sum
Mode
Mean
Median
Variance
End of Question 8
Question 9.
Which measure of central tendency is obtained by calculating the sum of values and
dividing this figure by the number of values there are in the data set?
Range
Mean
Sum
Median
Mode
Dispersion
End of Question 9
Question 10.
Which measure is the most unreliable indicator of central tendency if data are
skewed?
Mean
Mode
Distribution
Median
Range
None of these
End of Question 10
Question 11.
Calculating the difference between the largest and smallest figure produces which
figure?
Variance
Mode
Mean
Median
Range
Sum
End of Question 11
Question 12.
Which two measures use the mean as a baseline and identify the extent to which
scores differ from this?
End of Question 12
Question 13.
What does a large standard deviation suggest?
End of Question 13
Question 14.
Which percentage of scores falls within 1 standard deviation from the mean?
50%
68%
30%
99.7%
95%
9%
End of Question 14
Question 15.
What percentage of scores falls within three standard deviations from the mean?
95%
68%
15%
99.7%
50%
None of these
Chapter 5: Multiple Choice Questions
Try the multiple choice questions below to test your knowledge of this Chapter.
Once you have completed the test, click on 'Submit Answers' to get your results.
Question 1.
By what alternative name is Pearson's Correlation Analysis also known?
None of these
Mann-Whitney U Test
Pearson's Product-Moment
Spearman's Correlation Analysis
Kruskal-Wallis Correlational Analysis
Chi-Squared Product-Moment
End of Question 1
Question 2.
What does a Pearson's product-moment allow you to identify?
End of Question 2
Question 3.
What type of data is required for a Pearson's analysis which does not include a
dichotomous variable?
Nominal or ordinal
Categorical or ratio
Ordinal or interval
Ratio or nominal
Interval or ratio
Interval or nominal
End of Question 3
Question 4.
What type of relationships does a Pearson's product-moment assess?
End of Question 4
Question 5.
What must data be in order for a Pearson's product-moment to be conducted?
Parametric
All of these
Homoscedasticity
Free from outliers
Homogeneity of variance
Normally distributed
End of Question 5
Question 6.
Homoscedasticity can be checked using which type of graph?
Histogram
Pie chart
Scatter graph
Bar chart
Box-plot
Line graph
End of Question 6
Question 7.
A bell shaped curve to a scatter graph would suggest what?
None of these
There would be a non-linear relationship and a Pearson's product-moment
should be used
There would be a non-linear relationship and a Pearson's product-moment
should not be used
There is an outlier but a Pearson's product moment can still be used
There would be a linear relationship and a Pearson's product-moment should
be used
There would be a linear relationship but a Pearson's product-moment should
not be used
End of Question 7
Question 8.
If all points cluster in an ascending line this would suggest what?
End of Question 8
Question 9.
If most points depict a dispersed descending line this would suggest what?
End of Question 9
Question 10.
How should a significance level of 0.000 ideally be reported in APA format?
P > 0.01
P < 0.05
P > 0.05
0.000
P < 0.5
P < 0.001
End of Question 10
Question 11.
A Pearson test statistic of .876 with a significance level of P < 0.01 would suggest
what?
End of Question 11
Question 12.
When reporting a Pearson's product-moment in APA format, what letter do you use
to indicate which test you used?
This would be
This would be N
This would be F
This would be r
This would be P
This would be x
End of Question 12
Question 13.
What correlation can you use if your data do not meet the assumptions of normal
distribution?
Mixed ANOVA
Paired samples t-test
ANCOVA
Independent samples t-test
Spearman's rho
Chi-square
End of Question 13
Question 14.
A Spearman's test statistic of -.207 with a p = 0.057 would suggest what?
End of Question 14
Question 15.
When reporting a Spearman's Rho in APA format, what letter do you use to indicate
which test you used?
This would be
This would be P
This would be F
This would be x
This would be r
This would be N
End of Question 15
Question 16.
What does a partial correlation analysis allow you to do which Pearson's and/or
Spearman analyses do not?