Description
PLEASE COMPLETE THE ANOVA OUPUTS, AND PROVIDE ME WITH A WRITEUP (TEMPLATE FOR WRITE-UP PROVIDED)
Unformatted Attachment Preview
1
1-Factor ANOVA Overview
Between-Subjects ANOVA
*these analyses are from the dataset “Demo 1-Factor BS ANOVA.”
Cora and Talia are interested in whether subtle facial expressions affect perceptions of
threat. To assess this, they obtain photos of incarcerated white males from the state of
Florida corrections website and manipulate the faces so they are either scowling, smiling,
or neutral, and then randomly assign participants to either the scowling, smiling, or neutral
condition, and has participants rate how threatening they find each face.
The faces are below.
Scowl
Neutral
Smile
Face
Groups Variable: Condition: Facial Expression Type: Scowl, Neutral, or Smile (FaceExp)
Continuous Variable: Perceptions of threat (threat)
2
Example Point & Click Analyses
Running the Univariate Analysis of Variance (ANOVA):
Analyze → General Linear Model → Univariate
Enter your continuous variable into the “Dependent Variable” box and your groups variable
into the “Fixed Factor” box:
3
Next select “EM Means.” Select any variables you want the view the means for and drag
into the box. You can also click “compare main effects,” which is one way to perform posthoc mean comparisons. Another place to request post hoc tests is under “post hoc.”
Next select “Options.” You can request measures of effect sizes and descriptives (M, SD)
here.
4
Example Output
The M, SD, n on the
continuous variable
for each condition
The BS
ANOVA
The main effect of the groups variable
(FaceExp) on the continuous variable
(threat):
F(2, 106) = 4.51, p = .01, η2 = .08.
5
The M, SE on the
continuous
variable for each
condition
Pairwise
comparisons
comparing each
level of the groups
variable
“FaceExp” against
each other level.
You can
ignore this.
6
Within-Subjects ANOVA
*these analyses are from the dataset “Demo 1-Factor WS ANOVA.”
Pirlott and Neuberg (2014) were interested in whether heterosexual women differentially
expressed prejudice toward heterosexual, bisexual, and gay male targets. To examine this,
heterosexual women rated how negative they felt (continuous variable) towards
heterosexual male, bisexual male, and gay male targets, using a within-subjects design.
Groups Variable: Condition: Sexual Orientation of Target: heterosexual, bisexual, gay
Continuous Variable: Negativity
Note that because Ps rated their negativity toward the three groups, the groups variable
(heterosexual, bisexual, gay) is connected to the continuous variable ratings (negativity).
Thus, Ps have three ratings of negativity: HM_Neg (negativity toward hetero men), BM_Neg
(negativity toward bi men), and GM_Neg (negativity toward gay men).
Groups Variable: Target Sexual Orientation
Level 1: HM: Heterosexual Male Targets
Level 2: BM: Bisexual Male Targets
Level 3: GM: Gay Male Targets
Continuous Variable: Negativity
7
Example Point & Click Analyses
Running the WS ANOVA: Analyze → General Linear Model → Repeated Measures
SPSS automatically labels your groups variable as “factor1.”
You can rename your groups variable so it’s more meaningful for you.
Then indicate the number of levels or conditions in the groups variable.
Then click add.
8
Next drag the three (or more) levels of the group ratings into the Within-Subjects Variables
box. Note how the numbers 1, 2, 3 in the variable name correspond with the levels in the
variable window
Next select “EM Means.” Select any variables you want the view the means for and drag
into the box. You can also click “compare main effects,” which is one way to perform posthoc mean comparisons. Another place to request post hoc tests is under “post hoc.”
9
Next select “Options.” You can also request measures of effect sizes and descriptives (M,
SD) here.
10
Example Output
The M, SD for each level
(1, 2, 3) in the groups
variable “TOrient” on the
continuous variable
You can
ignore these.
Main Effect of
Target
Orientation
11
The WS
ANOVA
The main effect of the TOrient on the negativity:
F(2, 154) = 5.91, p = .003, η2 = .07.
You can
ignore this.
The BS ANOVA. We
didn’t include any
BS variables.
12
The M, SE for each
level (1, 2, 3) in the
groups variable
“TOrient”
Pairwise
comparisons
comparing each
level of the groups
variable “TOrient”
against each other
level.
You can
ignore this.
13
APA Write-Up Information
Write-Up Template:
− To test whether [continuous variable] differed as a function of [groups variable], I
performed a [#levels of groups variable] ([Groups Variable]: [Condition/Level 1],
[Condition/Level 2], [Condition/Level 3], …) between-subjects ANOVA on
[continuous variable].
− Results were [significant/not significant], [report F-Statistic, including effect size].
− Include M, SD of each group/condition.
− Report planned comparison or post hoc tests to test whether the continuous
variable differed between each pair of conditions. Include p-value for each
comparison and the direction of the difference (or non-difference) for each
comparison.
− Discuss whether results support or don’t support your hypothesis and whether you
reject or fail to reject the null hypothesis.
14
One-Factor ANOVA Lab Example Write-Up
Angela G. Pirlott
Department of Psychology, Saint Xavier University
PSYCH 300: Statistics for the Social Sciences
Professor Angela G. Pirlott
[due date]
15
One-Factor ANOVA Lab Example Write-Up
Between-Subjects ANOVA
Do perceptions of a person as threatening depend on the person’s facial expression, i.e., if
the person is smiling, scowling, or neutral? I hypothesize that, because angry faces might signify
intent to cause harm, perceptions of threat differ as a function of facial expression, such that
scowling faces are perceived as more threatening than smiling or neutral faces. The null
hypothesis is that perceptions of threat do not differ as a function of facial expression.
Within-Subjects ANOVA
Does negativity differ as a function of a target’s sexual orientation? I hypothesize that
women dislike gay and bisexual men relative to heterosexual men, because in general, people are
prejudiced against LGB people. The null hypothesis is that women’s negativity does not differ as
a function of sexual orientation target.
Results
Between-Subjects ANOVA
To test whether perceptions of threat differed as a function of criminal’s facial expression
type, we ran a 3 (Facial Expression: Scowl, Neutral, Smile) between-subjects ANOVA on
perceptions of threat. The ANOVA was significant: F(2, 106) = 4.51, p = .01, η2 =.08, therefore,
we reject the null hypothesis. To determine which conditions significantly differed, we ran Least
Significant Differences planned comparisons. Scowling faces were perceived as significantly
more threatening (M = 4.71, SD = 1.01) than smiling faces (M = 4.03, SD = 1.18), p = .01.
Scowling faces were also perceived as significantly more threatening than neutral faces (M =
4.00, SD = 1.21), p = .01. Perceptions of threat did not significantly differ if the criminal was
smiling or neutral, p = .90.
16
Within-Subjects ANOVA
To test whether heterosexual women expressed prejudice differentially toward gay,
bisexual, and heterosexual male targets, we ran a 3 (Target Sexual Orientation: Heterosexual,
Bisexual, Gay) within-subjects ANOVA on general negativity. The ANOVA was significant:
F(2, 154) = 5.19, p = .003, η2 = .07, therefore, we reject the null hypothesis. To determine
exactly which groups differed from one another, we performed Least Significant Differences
planned comparisons. Women were significantly more prejudiced towards bisexual men (M =
2.63, SD = 2.46) than gay men (M =1.91, SD = 1.79), p < .001. They were also significantly
more prejudiced toward bisexual men then heterosexual men (M = 1.85, SD = 1.35), p = .01.
They were no more prejudiced toward gay men than toward heterosexual men, p = .79.
Purchase answer to see full
attachment
