Transcript Science Jeopardy

T Test Jeopardy
Name that
Statistical Test
Let’s do
research!
Fun with
SPSS
Right or
Wrong?
Hodgepodge
100
100
100
100
100
200
200
200
200
200
300
300
300
300
300
400
400
400
400
400
500
500
500
500
500
Final Jeopardy
FINAL JEOPARDY
• ANSWER: In SPSS, this number
represents the exact probability of
obtaining the result you did merely by
chance (AKA if the H0 is assumed to be
true)
• QUESTION: What is the significance
level or p-value reported under the
column Sig.(2-tailed) in SPSS?
Answer
Question
Name that Statistical Test
100
• ANSWER: An instructor tries a new method
of teaching statistics using humor in one
section and the old method in another. This
is the type of research design she is using,
and this is the type of t test she would do to
evaluate her research question.
• QUESTION: What is a between-subjects
design and an independent-samples t test?
Answer
Question
Name that Statistical Test
200
• ANSWER: A researcher wants to know if
his sample’s political orietation (rated by
Ps on a 9-point Likert scale from “very
conservative” to “very liberal”) is
significantly different from moderate (5
on the scale).
• QUESTION: What is a single-group
design and a one-sample t test?
Answer
Question
Name that Statistical Test
300
• ANSWER: A clinical psychologist compares
her clients general level of functioning
(assessed by the Global Assessment of
Functioning [GAF] scale; Axis V of the
DSM-IV-TR, 2000) before therapy and then
again after 10 sessions.
• QUESTION: What is a repeated measures
design and a paired-samples t test?
Answer
Question
Name that Statistical Test
400
• ANSWER: A cognitive psychologist compares
memory for nonsense syllables for participants
who are either instructed in the use of a
mnemonic strategy or who are not given any
strategy. He is only interested in if the
mnemonic strategy improves memory.
• QUESTION: What is a between-subjects
design and a one-tailed independent-samples t
test?
Answer
Question
Name that Statistical Test
500
• ANSWER: A school psychologist matches
students who are similar to one another on
math ability and then assess performance of
the two groups of students after one group
participates in a special math program.
• QUESTION: What is a matched-samples
research design and a paired-samples t test?
Answer
Question
Let’s do research!
100
• ANSWER: You find a significant result,
but your not sure whether the difference
is big enough to be important so you
decide to calculate this.
• QUESTION: What is effect size or
Cohen’s d?
Answer
Question
Let’s do research!
200
• ANSWER: A friend of your in methods
calculate’s Cohen’s d and gets a result of
d = 2.75. He is not sure what this means,
so this is how you tell him to interpret it.
• QUESTION: What is Holy shit!! That’s a
huge effect!!?
Answer
Question
Let’s do research!
300
• ANSWER: This is when calculating
Cohen’s d is appropriate.
• QUESTION: What is when you are doing
a paired-samples or independent-samples
t test, and the difference is significant?
Answer
Question
Let’s do research!
400
• ANSWER: Oh no, Levene’s test is
significant!! This is what it means, and
this is what I should do now.
• QUESTION: What is conclude that the
homogeneity of variance assumption for
the independent-samples t test is violated
and report t value and p value for the
second row (“equal variances not
assumed”)?
Answer
Question
Let’s do research!
500
• ANSWER: In APA style, you should
always report all of these things.
• QUESTION: What are means, standard
deviations, degrees of freedom, obtained t
value, p value, and Cohen’s d when
appropriate?
Answer
Question
Fun with SPSS
100
• ANSWER: In SPSS, the probability that
you would use to evaluate significance
when α = .01, and you are performing a
one-tailed test.
• QUESTION: What is .01 * 2 = .02?
Answer
Question
Fun with SPSS
200
One-Sample Test
Tes t Value = .5
disbs
t
4.728
df
29
Sig. (2-tailed)
.000
Mean
Difference
.46874
95% Confidence
Interval of the
Difference
Lower
Upper
.2660
.6715
• ANSWER: In a one-sample t test, the
confidence interval for these data.
• QUESTION: What is CI: .77 ≤ µ ≤ 1.17?
Answer
Question
Fun with SPSS
300
• ANSWER: In an independent-samples t
test, the row from which you report
results when Levine’s test is not
significant.
• QUESTION: What is “equal variances
assumed”?
Answer
Question
Fun with SPSS
400
Independent Samples Test
Levene's Test for
Equality of Variances
F
I am active in one or
more student
organization on campus .
Equal variances
ass umed
Equal variances
not as sumed
Sig.
49.566
.000
t-tes t for Equality of Means
t
df
Sig. (2-tailed)
Mean
Difference
Std. Error
Difference
95% Confidence
Interval of the
Difference
Lower
Upper
3.025
34
.005
2.15584
.71279
.70728
3.60441
2.522
15.044
.023
2.15584
.85494
.33406
3.97763
• ANSWER: In an independent-samples t
test, the confidence interval for the SPSS
output reported above.
• QUESTION: What is CI: .33 ≤ µ ≤ 3.98?
Answer
Question
Fun with SPSS
500
Independent Samples Test
Levene's Test for
Equality of Variances
F
I am active in one or
more student
organization on campus .
Equal variances
ass umed
Equal variances
not as sumed
49.566
Sig.
.000
t-tes t for Equality of Means
t
df
Sig. (2-tailed)
Mean
Difference
Std. Error
Difference
95% Confidence
Interval of the
Difference
Lower
Upper
3.025
34
.043
2.15584
.71279
.70728
3.60441
2.522
15.044
.065
2.15584
.85494
.33406
3.97763
• ANSWER: The conclusion a researcher would
draw from these data.
• QUESTION: What is retain H0 and conclude
there is no significant difference between
students who prefer morning and afternoon
classes?
Answer
Question
Right or Wrong?
100
• ANSWER: The type of error that occurs
when I fail to reject a false null
hypothesis.
• QUESTION: What is Type II error?
Answer
Question
Right or Wrong?
200
• ANSWER: The type of error that occurs
when I reject a true null hypothesis.
• QUESTION: Type I error?
Answer
Question
Right or Wrong?
300
• ANSWER: The more of this you have,
the more likely it is that you will make a
correct decision about the null
hypothesis.
• QUESTION: What is power?
Answer
Question
Right or Wrong?
400
• ANSWER: As the probability of Type II
error decreases, this simultaneously
increases.
• QUESTION: What is power?
Answer
Question
Right or Wrong?
500
• ANSWER: When you increase α, the
“good news” is that the probability of this
kind of mistake goes down, but this
comes at a price because the probability
of this other kind of mistake
simultaneously increases.
• QUESTION: What is Type II error and
Type I error?
Answer
Question
Hodgepodge
100
• ANSWER: The only circumstance under
which using a one-tailed test is
appropriate.
• QUESTION: What is when an outcome
opposite to that you are predicting means
exactly the same thing to you as finding
there’s no difference?
Answer
Question
Hodgepodge
200
• ANSWER: In the independent-samples t
test, the assumption that the variances in
the two populations you are comparing
are equal.
• QUESTION: What is the homogeneity of
variance assumption?
Answer
Question
Hodgepodge
300
• ANSWER: Increasing sample size
decreases this, which then increases this.
• QUESTION: What is the standard error
of the estimate and power?
Answer
Question
Hodgepodge
400
• ANSWER: In an independent-samples t
test, the null and alternative hypotheses
when I’m predicting that anxiety patients
who get therapy will have less anxiety
than those that don’t.
• QUESTION: What is H0: µtherapy - µno
therapy ≥ 0 and HA: µtherapy - µno therapy < 0?
Answer
Question
Hodgepodge
500
• ANSWER: In an independent-samples t
test, this is the null hypothesis that
Levine’s test evaluates.
• QUESTION: What is the null hypothesis
that the variance in the two samples
being compared are equal?
Answer
Question