How To Write the Results and Discussion

1
How To Write the Results and Discussion Sections
2
(No Transcript)
3
(No Transcript)
4

• Results
• State that you are presenting the mean proportion
  of false
• recall in either a table or figure.

5
(No Transcript)
6
(No Transcript)
7

• Results
• State that you are presenting the mean proportion
  of false
• recall by list type and warning condition in
  either a table or
• figure.

8
FALSE MEMORY 10
Figure 1. The proportion of false recall as a
function of list type and warning status.
9
Results 1. State that you are presenting the mean
proportion of false recall by list type and
warning condition in either a table or figure.
2. State what type of analysis was conducted on
the proportion of false recall (i.e., a
two-factor mixed design ANOVA).
10
Results 1. State that you are presenting the mean
proportion of false recall by list type and
warning condition in either a table or figure.
2. State what type of analysis was conducted on
the proportion of false recall (i.e., a
two-factor mixed design ANOVA). 3. State the
results of this analysis A. a significant main
effect of list type, F(2, 144) 21.35, p lt
.001 B. a significant main effect of warning
status, F(1, 72) 9.74, p .003 C. a
marginally significant interaction between list
type and warning status, F(2, 144)
3.05, p .051
11
Results 4. Then say (in the same paragraph) that
paired-samples t-tests were conducted to
determine if there were significant differences
in the rate of false recall between list type
conditions.
12
Results 4. Then say (in the same paragraph) that
paired-samples t-tests were conducted to
determine if there were significant differences
in the rate of false recall between list type
conditions.
why do we do these t tests? need to indicate
why in paper
13
Results 4. Then say (in the same paragraph) that
paired-samples t-tests were conducted to
determine if there were significant differences
in the rate of false recall between list type
conditions.
Note that, for all post hoc tests, the Bonferroni
correction was applied.
14
Results 4. Then say (in the same paragraph) that
paired-samples t-tests were conducted to
determine if there were significant differences
in the rate of false recall between list type
conditions. Again when you report these results,
you must use complete sentences, e.g., "The
difference between the rate of false recall
produced by... was significant, t(71) 6.21, p
lt .001.") Here are the results of those
t-tests 1. semantic vs. phonological, t(71)
.85, p 1.0. 2. semantic vs. hybrid, t(71)
6.21, p lt .001 3. phonological vs. hybrid, t(71)
4.72, p lt .001.
15
Results 5. Then say (in the same paragraph) that
due to ? independent samples t-tests were
conducted comparing the rate of false recall
between warned and unwarned conditions for each
of the list types.
16
Results 5. Then say (in the same paragraph) that
due to the marginally significant interaction
between list type and warning status,
independent samples t-tests were conducted
comparing the rate of false recall between
warned and unwarned conditions for each of the
list types.
17
Results 5. Then say (in the same paragraph) that
due to the marginally significant interaction
between list type and warning status,
independent samples t-tests were conducted
comparing the rate of false recall between
warned and unwarned conditions for each of the
list types.
Here are the results of those t-tests (Again when
you report these results, you must use complete
sentences) 1. semantic, t(70) 2.36, p .06
indicate that this was marginally significant 2.
hybrid, t(70) 3.16, p .006 3. phonological,
t(70) .23, p 1.0   The Discussion section
will begin right after you finish reporting the
results of these t-tests.
18
FALCE MEMURY 11 Results
The proportion of false recall by condition is
presented in Figure 1. A multifactor ANOVA was
conducted on the proportion of false recall.
The results yielded a main effect of list
type, F (2, 352) 16.47, p lt .001, and a List
Type X Recall Modality interaction, F (2, 352)
7.55, p lt .001. Post hoc Boneferroni t-tests
indicated that the difference between the
proportion of false recall produced in the
written mode and the verbal mode differed only
in the pseudohomophone condition, t (178) 4.14,
p lt .001, both other ts lt 1. Discussion
The results are consistent with the idea
that one can source monitor the list context in
order to determine the plausibility of an
activated memory, and the plausibility of an
activated memory influences whether or not it
will be recalled. When a list consists entirely
of pseudohomophones, the critical lure will be
activated in memory. However, when attention is
directed toward orthography as in the written
recall mode, the activated lure may be rejected
as implausible because of the mismatch between
the orthography that is activated in memory and
the list context. In contrast, the other two list
types contain correctly spelled words.
Therefore, the form of the lure word that is
activated in memory remains plausible when
compared to the list
19
Discussion a) Begin by indicating if your
original hypotheses were supported or not. - if
not, offer explanation As predicted (replicated
Watson et al. 2003) - Hybrid lists produced a
higher rate of false recall than either
phonological lists or semantic lists which
produced similar rates of false recall. The
original hypothesis regarding warning was only
partially supported - Warning participants about
the critical lure reduced false recall
substantially in the hybrid and semantic
conditions but not the phonological
condition.   You will need to offer explanation
as to why warning did not reduce false recall in
the phonological condition. State the reason
here, and then indicate that you will elaborate
on this hypothesis later.
20
Discussion b) Explain implications of the
research and describe the findings in the context
of the extant literature - discuss list type
effect as it relates to the original findings of
Watson, Balota, and Roediger (2003). - indicate
that this effect is consistent with the source
monitoring hypothesis discussed in the
Introduction. you wont provide too much here as
you will have discussed it in more detail in the
Intro., and it is not the main focus of the
experiment.   - discuss warning effect as it
relates to findings in the literature - Gallo et
al. - discuss hypothesis regarding warning (i.e.,
that warning participants allows them to ID the
CL during study so that it can be avoided at
recall). - then hone in on the Neuschatz et al.
(2003) studies, and discuss their method,
results, and interpretation. Then tie it to our
study.
21
Discussion c) Discuss future research You
should propose a study that could be used to test
the idea that it may have been more difficult to
identify the critical lure in the phonological
condition than in the other conditions. When you
do propose this study, try to be fairly specific
about the conditions, measures, procedures,
predictions, and rationales.
22
Discussion d) Draw conclusions Focus on the
unique aspects of our study that are important.
23
Verbs tenses (see APA, 2.02 pp.
32-33) Results past tense (False recall
increased as a function of) Use present tense
to interpret the results and present conclusions
(e.g., in Discussion section) the results of
Experiment 2 indicate
24
Class Project Fall 2011 why didnt the
warning reduce false recall for phonological
lists? Prediction false recall would be
reduced in the warning condition Rationale
participants that are warned will ID the critical
lure and select against it at recall
25
Neuschatz, Benoit, Payne (2003) Effective
warnings in the Deese-Roediger-McDermott
false-memory paradigm The role of
identifiability basic hypothesis the more
identifiable the lure, the more effective the
warning will be
26
Neuschatz, Benoit, Payne (2003) Effective
warnings in the Deese-Roediger-McDermott
false-memory paradigm The role of
identifiability Norming study to determine
identifiability HI lists gt 66 of time critical
lure IDed (mean 80) - false recognition
.77 Stadler et al. (1999) LI lists lt 41 of
the time critical lure IDed (mean 29) -
false recognition .76 Stadler et al. (1999)
27
Neuschatz, Benoit, Payne (2003) Effective
warnings in the Deese-Roediger-McDermott
false-memory paradigm The role of
identifiability Design The design was mixed and
consisted of two IVs. Warning Status (warned, not
warned) was manipulated between subjects, and
Lure Identifiability (high, low) was manipulated
within subjects. Proportion falsely recognized
served as the dependent variable.
28
Neuschatz, Benoit, Payne (2003) Effective
warnings in the Deese-Roediger-McDermott
false-memory paradigm The role of
identifiability Design The design was mixed and
consisted of two IVs. Warning Status (warned, not
warned) was manipulated between subjects, and
Lure Identifiability (high, low) was manipulated
within subjects. Proportion falsely recognized
served as the dependent variable.