STATISTICS FOR PSYCHOLOGISTS

PART IV:
STATISTICS IN APA STYLE


Section Abstract: This section describes basic rules for presenting statistical results in APA style.
All rules come from the newest APA style manual. Specific examples of mini Results summaries
(and data tables) are provided, using the analyses in the previous section of this project.

Keywords: APA style, Results sections, statistical interpretation
Section Updated: December 2012













This document is part of an online statistics textbook.
Access to the complete textbook, along with licensing information, is available online:
http://www4.uwsp.edu/psych/cw/statistics/textbook.htm


Table of Contents for This Section

GENERAL RULES FOR APA STYLE RESULTS SECTIONS ..................................................................................... 2
EXAMPLES OF APA STYLE ...................................................................................................................... 3
SUMMARY OF PARAMETRIC STATISTICS ..................................................................................................... 5

Page 2 of 6
GENERAL RULES FOR APA STYLE RESULTS SECTIONS


Overview

The APA manual describes appropriate strategies for presenting statistical information. These guidelines were established
to provide basic minimal standards and to provide some uniformity across studies.


Using a Sufficient Set of Statistics

Information to Include: Significance testing is but a starting point and that additional reporting elements such as effect
sizes, confidence intervals, and extensive description are needed to convey the most complete meaning of the results (p.
33).

1. Descriptive statistics are essential and such a set usually includes at least the following: the per-cell sample sizes; the
observed cell means (or frequencies of cases in each category for a categorical variable); and the cell standard
deviations (p. 33).
2. For statistical significance tests, include the obtained magnitude or value of the test statistics, the degrees of
freedom, the probability of obtaining a value as extreme as or more extreme than the one obtained (the exact p
value) (p. 34).
3. When possible, confidence intervals should be emphasized. The inclusion of confidence intervals (for estimates of
parameters, for functions of parameters such as differences in means, and for effect sizes) can be an extremely
effective way of reporting results (p. 34).
4. For the reader to appreciate the magnitude or importance of a studys findings, it is almost always necessary to
include some measure of effect size (p. 34). These can be in the original (raw) units or in a standardized metric.


Information in Text versus Data Displays: Statistical and mathematical copy can be presented in text, in tables, and in
figures. . . Select the mode of presentation that optimizes understanding of the data by the reader (p. 116).

Generally speaking, the more data you have, the more likely it is that they should be presented in a table or figure. If
you need to present 4 to 20 numbers, first consider a well-prepared table (p. 116).
If you present descriptive statistics in a table or figure, you do not need to repeat them in text, although you should
(a) mention the table in which the statistics can be found and (b) emphasize particular data in the narrative when
they help in interpretation (p. 117).
As a result, it is necessary that the text include a description of the variable(s) under study and a description of the
statistical procedures used. The text often includes a description of whether the results support the hypotheses.












All quotations pertaining to reporting results are taken from: American Psychological Association. (2010).
Publication manual of the American Psychological Association (6
th
Ed.). Washington, DC: APA.

Page 3 of 6
EXAMPLES OF APA STYLE RESULTS IN THE TEXT


Descriptive Statistics: The purpose of the descriptive statistics is to provide the reader with an idea about the basic
elements of the group(s) being studied. Note that this also forms the basis of the in-text presentation of descriptive
statistics for the inferential analyses below.

On the quiz, the nine students had a mean score of 7.000 (SD = 1.225). Scores
of 6.000, 7.000, and 8.000 represented the 25th, 50th, and 75th percentiles,
respectively.


Correlations: Correlations provide a measure of statistical relationship between two variables. Note that correlations can
be tested for statistical significance (and that this information should be summarized if it is available and of interest).

For the nine students, the scores on the first quiz (M = 7.000, SD = 1.225) and
the first exam (M = 80.889, SD = 6.900) were strongly and significantly
correlated, r(8) = .695, p = .038.


One Sample t Test: In this case, a sample mean has been compared to a user-specified test value (or a population mean).
Thus, the summary and the inferential statistics focus on that difference.

A one sample t test showed that the difference in quiz scores between the current
sample (N = 9, M = 7.000, SD = 1.225) and the hypothesized value (6.000) were
statistically significant, t(8) = 2.449, p = .040, 95% CI [0.059, 1.941], d =
0.816.


Independent Samples t Test: For this analysis, the emphasis is on comparing the means from two groups. Here again the
summary and the inferential statistics focus on the difference.

An independent sample t test showed that the difference in quiz scores between
the control group (N = 4, M = 6.000, SD = 0.817) and the experimental group (N
= 4, M = 8.000, SD = .817) were statistically significant, t(6) = -3.464, p =
.013, 95% CI [-3.413, -0.587], d = -2.449.


One Way ANOVA: The ANOVA provides an omnibus test of the differences across multiple groups. Because the ANOVA
tests the overall differences among the groups, the text discusses the differences only in general.

A one way ANOVA showed that the difference in quiz scores between the control
group (N = 3, M = 4.000, SD = 1.000), the first experimental group (N = 3, M =
8.000, SD = 1.000), and the second experimental group (N = 3, M = 9.000, SD =
1.000) were statistically significant, F(2,6) = 21.000, p = .002,
2
= .875.



Page 4 of 6
One Way ANOVA with Post Hoc Tests: Post hoc tests build on the ANOVA results and provide a more focused
comparison among the groups. Notice that the post hoc summary duplicates the presentation of the omnibus ANOVA
statistics.

A one way ANOVA showed that the difference in quiz scores between the control
group (N = 3, M = 4.000, SD = 1.000), the first experimental group (N = 3, M =
8.000, SD = 1.000), and the second experimental group (N = 3, M = 9.000, SD =
1.000) were statistically significant, F(2,6) = 21.000, p = .002,
2
= .875.
Tukeys HSD tests showed that both experimental groups scored statistically
significantly higher than the control group. However, the two experimental groups
did not differ significantly.


Repeated Measures ANOVA: The RMD ANOVA tests for overall differences across the repeated measures. As such, its
summary parallels that of the One Way ANOVA.

A repeated measures ANOVA showed that, for the five people, the difference in
quiz scores between the first time point (M = 6.400, SD = 1.140) and second time
point (M = 7.800, SD = 0.837) were statistically significant, F(1,4) = 32.667,
p = .005, partial
2
= .875.


Factorial ANOVA: The factorial ANOVA provides statistics for all of the main effects and interactions in a factorial design.
Each effect would be summarized in a style analogous to a One Way ANOVA.

A 2 (Factor A) x 2 (Factor B) ANOVA was conducted on the quiz scores. Neither
Factor A, F(1,8) = 0.000, p = 1.000, partial
2
= .000, nor Factor B, F(1,8) =
.750, p = .412, partial
2
= .086, had a statistically significant impact on quiz
scores. However, the interaction was statistically significant, F(1,8) = 6.750,
p = .032, partial
2
= .458. The descriptive statistics for these analyses are
presented in Table 1.
Page 5 of 6
EXAMPLES OF APA STYLE RESULTS IN TABLES

Page 6 of 6
SUMMARY OF PARAMETRIC STATISTICS


Statistic What Its Purpose Is How To Report It What It Indicates
Descriptive Statistics
Mean
To provide an estimate of the
population from which the sample was
selected.
M = _____
Indicates the center point of the
distribution and serves as the
reference point for nearly all
other statistics.
Standard
Deviation
To provide an estimate of the amount
of variability/dispersion in the
distribution of population scores.
SD = _____
Indicates the variability of scores
around their respective mean.
Zero indicates no variability.
Measures of Effect Size
Cohens d
To provide a standardized measure of
an effect (defined as the difference
between two means).
d = _____.
Indicates the size of the
treatment effect relative to the
within-group variability of scores.
Correlation
To provide a measure of the
association between two variables
measured in a sample.
r(df) = _____
Indicates the strength of the
relationship between two
variables and can range from 1
to +1.
Eta-Squared
To provide a standardized measure of
an effect (defined as the relationship
between two variables).

2
= _____.
Indicates the proportion of
variance in the dependent
variable accounted for by the
independent variable.
Confidence Intervals
CI for a Mean
To provide an interval estimate of the
population mean. Can be derived from
both the z and t distributions.
____% CI [ ___, ___ ]
Indicates that there is the given
probability that the interval
specified covers the true
population mean.
CI for a Mean
Difference
To provide an interval estimate of the
population mean difference. Can be
derived from both the z and t
distributions.
____% CI [ ___, ___ ]
Indicates that there is the given
probability that the interval
specified covers the true
population mean difference.
Significance Tests
One Sample t
Test
To compare a single sample mean to a
population mean when the population
standard deviation is not known
t(df) = ____, p = ____.
A small probability is obtained
when the statistic is sufficiently
large, indicating that the two
means significantly differ from
each other.
Independent
Samples t Test
To compare two sample means when
the samples are from a single-factor
between-subjects design.
Related Samples
t Test
To compare two sample means when
the samples are from a single-factor
within-subjects design.
One-Way
ANOVA
To compare two or more sample
means when the means are from a
single-factor between-subjects design.
F(df
1
,df
2
) = ___, p = ___.
A small probability is obtained
when the statistic is sufficiently
large, indicating that the set of
means differ significantly from
each other.
Repeated
Measures
ANOVA
To compare two or more sample
means when the means are from a
single-factor within-subjects design.
Factorial
ANOVA
To compare four or more groups
defined by a multiple variables in a
factorial research design.

Note. Many of the statistics from each of the categories are frequently and perhaps often appropriately presented in
tables or figures rather than in the text.