PSYC 515
Page 1 of 5
HELPFUL HINTS FOR SUCCESS IN PSYC 515
Index of Topics:
How to Actively Practice Concepts
Overview and Strategy for Quizzes
Writing Results in APA – Hints and Examples (with statistical notations)
Statistical Decision Flow Chart (helps you select the most appropriate test AND graph)
How to Actively Practice Concepts
Before completing any assignment, read the module’s assigned readings and watch the presentations. Complete
the Critical Thinking Checks embedded within the “Research Methods and Statistics” text. For extra practice,
each chapter of the “Research Methods and Statistics” text has a review section that contains Chapter Exercises
and a Chapter Review. The answers are provided to the odd numbered “Chapter Exercises” and for every
question in the “Chapter Review”! The Chapter Exercises gives you a great opportunity to practice hand
calculations and SPSS similar to our homework assignments. The Chapter Review is more conceptual, and is
similar to our quizzes.
The e-book “A Simple Guide to IBM SPSS Statistics” (Parts I and II) is for helping with SPSS. You should work
through their Sample Problems using SPSS yourself for additional practice (your output should match that in the
e-book). Finally, the Publication Manual is required in the course – use it as a guide for all APA style formatting
questions (note it doesn’t give specific examples of statistical notation write-ups – those are only provided in the
powerpoint lectures!).
Overview and Strategy for Quizzes
Quizzes focus on concepts and theory whereas homework requires use of SPSS. There is NO use of SPSS for your
exams, although there are some hand-calculations. You have 90 minutes to complete 40 multiple choice
questions. Make sure you’ve given yourself ample time in your schedule so that you are not rushed. You can
have as many
owsers open as you want, so feel free to open all presentations, e-book chapters, et cet that
were particularly helpful before beginning the exam. Read all answer options. Some times “all of these options”
or “none of these options” is the co
ect answer. If you select the first “right” answer you see, you may miss
these! For questions that are particularly difficult, answer to the best of your ability and record that question
number. Revisit if you have time at the end of your quiz.
PSYC 515
Page 2 of 5
Example Results (with statistical notations) in APA Format for PSYC 515
Helpful Hints Regardless of Test:
• If the “sig” column in SPSS is a number equal to or higher than .05 that means you fail to reject the null
hypothesis and conclude it was NOT significant.
• If the “sig” column in SPSS is less than .05 (e.g., .048, .03, .001) that means you REJECT the null hypothesis
and conclude there is a significant effect. If df = 1 then just look at the means to interpret HOW it is
significant (if co
elation, look to see if it is positive or negative to interpret). If df > 1 then you need to
conduct the appropriate post hoc analysis to interpret it more fully.
• **Always report EXACT p value from SPSS output unless SPSS states .000, then you write p<.001**
Independent samples t-test – how to run this test in SPSS is shown in Module 1 as Review
• r2 =
??
??+??
where r2 = .01 small, r2 = .09 medium, r2 = .25 large
An independent samples t-test found a statistically significant difference in the amount of money donated based
on how child hunger is described t (18) = -2.383, p = .028, r2 = 0.24, 95% CI [-21.450, -1.350] (two-tailed),
ejecting the null hypothesis. The group given facts about child hunger donated significantly less money (M =
19.10; S = 7.06) than the group described an identifiable victim (M = 30.50; S = 13.38).
Co
elated groups t-test – how to run this test in SPSS is shown in Module 1 as Review
• r2 =
t2
t2 + df
A co
elated groups t test examined whether participating in sports will positively influence self-esteem in girls.
It was statistically significant t (5) = -6.71, p <.001, r2 = .90, 95% CI [-.93, -2.07]. The null hypothesis is rejected. As
shown in Figure 1, sports participation positively influences self-esteem in girls.
Pearson’s r co
elation – how to run this test in SPSS is shown in Module 1 as Review
A Pearson’s r co
elation revealed a significant relationship between the self-concept of intimate relationships
and friends, r (78) = 0.552, p <.001 (two tailed). The null hypothesis is rejected; 30.47% of the variation in
intimate relationships is accounted for by friends.
Simple linear regression analysis – how to run this test in SPSS is shown in Module 1 as Review
A linear regression analysis was conducted to evaluate the prediction for the self concept of friends given the
self concept of intimate relationships and was found to be significant F (1,78) = 34.231, p <.001. The regression
equation for predicting the self concept of friends is ?′ = .617X XXXXXXXXXXThe co
elation between self concepts
of friends and intimate relationships is XXXXXXXXXXApproximately 30.47% of the variance in the self concept of
friends was accounted for by its linear relationship with the self concept of intimate relationships.
Randomized ANOVA – how to run this test in SPSS is shown in Module 1
• Ƞ2 =
?????????
???????
A One-Way ANOVA was conducted to examine whether a preceding situation (watching a video of helping
ehavior, seeing first-hand someone help another person, or a neutral control condition) influenced the number
of helping behaviors expressed by people. The null hypothesis was rejected - there is a difference between the
control, video, and live conditions in number of helping behaviors F(2, 21) = 4.993, p = .017, Ƞ2 = XXXXXXXXXXTukey
post hoc analyses revealed that seeing someone first-hand help another person resulted in significantly more
helpful behaviors than being in the neutral control condition (p=.013; see Figure 1). No other post hoc
comparisons were significant (p>.05).
PSYC 515
Page 3 of 5
Repeated measures ANOVA – how to run this test in SPSS is shown in Module 1
A One-Way Repeated Measures ANOVA was conducted to examine whether a preceding situation (watching a
video of helping behavior, seeing first-hand someone help another person, or a neutral control condition)
influenced the number of helping behaviors expressed by people. The null hypothesis was rejected - there is a
difference between the control, video, and live conditions in number of helping behaviors F(2, 14) = 4.20, p =
.037, Ƞp2 = .375. Post hoc pairwise comparisons revealed that the control group had significantly fewer helping
ehaviors than the live group (p=.015). No other comparison was significant (p>.05) (see Figure 1).
Factorial Design – how to run this test in SPSS is shown in Module 2 *note you must report each main effect (one
per variable) and all interactions (if you have two variables you will only have one interaction). If anything is
significant with df > 1 you will need to conduct an appropriate post hoc analysis, as shown on slide 28 in the
Factorial Designs in SPSS presentation in Week 2.
A 2 x 3 ANOVA was conducted to assess the effects of sex (male/female) and diet type (calorie restriction,
ketogenic, and vegan) on weight loss. There was no main effect of sex [F(1,54) = 0.054, p=.816, Ƞp2=.001] and no
main effect of diet [F(2,54) = 2.447, p=.096, Ƞp2=.083 ]. The interaction of sex x diet was significant [F(2,54) =
3.585, p=.035, Ƞp2=.117]. Post hoc analyses were conducted using independent samples t-tests for each diet
type and revealed that females lost more weight (M=13.32, SD=3.11) than males (M=8.55, SD=5.36) on the
vegan diet (p=.026) (see Figure 1). Amount of weight lost did not differ by sex for the low calorie diet (p=.798) or
the ketogenic diet (p= .141).
Chi Square Test of Independence – how to run this test in SPSS is shown in Module 4
• ? = √
?2
?
A Chi Square test of independence revealed a significant effect of background influencing willingness to use
mental health services ?2(1, N=80) = 0.990, p = .320, ? = .34. As shown in the Figure 3, female college students
from rural areas were less likely to be willing to use mental health services than students from subu
s or u
an
areas.
For any other test, refer to the examples in the presentations.
PSYC 515
Page 4 of 5
STATISTICAL TEST SELECTION GUIDE for PSYC 515
• Module for topic is in parentheses
• the *typically* prefe
ed graph type is shown for each test
IV = independent variable; DV = dependent variable
BS = between subjects; WS = within subjects
R
EL
A
TI
O
N
SH
IP
S
(M
o
d
u
le
1
R
ev
ie
w
f
o
m
P
SY
C
5
1
0
)
Use scale of measurement of both
variables to determine which
co
elation test
Pearson (scatterplot)
oth scale
Spearman (scatterplot or bar graph, depending
on # of levels)
oth ordinal
Point Biserial (bar graph)
one nominal dichotomous;
one scale
Phi co
elation
(clustered bar graph)
oth nominal and dichotomous
also want to PREDICT
Simple Regression
(scatterplot with regression line)
(what you try to predict is the DV)
Differences
when DV is
NOMINAL
(module 4)
compare ONE observed nominal
data with provided, previous
knowledge / theory Chi Square Goodness of Fit
(bar graph)
determine whether two
observed nominal variables are
associated with one anothe
Chi Square Test of Independence
(clustered bar graph)
PSYC 515
Page 5 of 5
"DIFFERENCES" / "EFFECTS" when DV is SCALE
*Nonparametric test equivalent used if DV is ordinal or data fails to meet assumptions