1/68
Looks like no tags are added yet.
Who needs to submit to an IRB?
Study constitutes research
on human subjects
Types of IRB reviews
Exempt, expedited, full board
Exempt
Reviewed by chair and 1 member. to qualify study must be in an educational setting, be a survey, test, interview, observation, or benign behavioral intervention
Expedited
needs slightly more review. Reviewed by chair or vice chair and 2 members
Full board
All members look at it, and researcher comes in to answer questions
Blanket protocol
Davidson specific - professor submits for the whole class to speed up the process.
attachments in IRB submission
Researcher project description with complete materials, consent form, debriefing, human subjects certs, faculty sponsor assurance form
Research project description
purpose of study and hypothesis, description of procedure, exact materials. Should all be written in lay person language
Mundane realism
trying to replicate real life for experimental purposes
expiremental realism
Trying to make the study involving typically has more control and fewer confounds.
psychological realism
trying to make the psychological expirience as similar to real life as possible
construct valididty
how well the test measures what we want it to
face valididty
how well does the scale appear to measure the concept. High face validity could influence demand characteristics and will affect deception
convergent validity
does the measure align with other measures of similar concepts (could be positive or negative)
discriminant validity
is the measure distinct from other measures
internal validity
is the change in the DV only connected to change in the IV - no confounds or uncontrolled IDVs
external validity
do the results of the study generalize to other settings
threats to internal validity
M- maturation: people change overtime
R- regression: extremes move back toward mean regardless of what we do
S- selection: non random assignment causing a confound
M- mortality: attrition / we lose participants
I - instrumentation: how well your instrument works
T - Testing: how well your test works
H: history - worldly context outside of the study
Realtionship between internal and external validity

Psychometric Reliability
internal scale reliability
test retest
split half
Experimental Reliability
replication
Descriptive statistics
Frequency, Central tendency (mean, median, mode), Variability (variance, st. dev, range)
Standardized scores (z distribution)
mean will = 0, st dev = 1
Inferential statistics
purpose is to infer comparisons between sample and population
Steps of Hypothesis testing
state hypothesis (null and alternative)
draw a random sample
compare sample and prediction
reject null or fail to reject
Type 1 Error
reject a true null hypothesis - false positive
Type 2 Error
Failed to reject a false null - False negative
Independent samples t test
compares mean to difference expected by chance
repeated measures t test
compares average observed difference between time 1 and time 2 to difference expected by chance
one way anova
Are the means equal across three measures or groups
repeated measures anova
variance between treatments compared to error
factorial anova
tests main effects - for three or more IDVs
Mixed anova
for at least 1 between subjecst variable and 1 within subjects variable
Post hoc tests of anova
shows you which conditions are influencing a significant relationship
Concerns around post hoc test
Running more post hoc tests increases likely hood of committing a type one error. Because of this, correction tests are used (boneferroni, tukeys HSD, LSD)
Factors influencing hypothesis testing
size of mean difference, variability of scores, number fo scores
ways to increase power
larger sample size, tight experimental control, strong manipulation
statistical assumptions
independence of observations and population is evenly distributed
Correlation
measures strength and direction of a relationship
regression
predicts y from x by identifying the slope of the best-fitting line.
strict significance testing
.05 is an arbitrary number, and the field is trying to move away from determining success just from it
qualitative data
anything that doesn’t start in numerical form
objective coding
no interpretation needed. Class year or number of words used
subjective coding
Not as easy as counting, some quality is being assessed. Narrative coding, thematic analysis, concept coding
Steps of Thematic Coding
open coding, axial coding, selective coding
Open coding
First pass through material, looking for info in light of research question and create general themes
axial coding
A second pass that occurs multiple times. Read through after establishing code and adjust acordingly
Selective coding
Identifying specific cases of each code
content analysis
can be objective or subjective, but a very common analysis on secondary data
reasons for using Archival Data
limited access to participants, limits to funding, question is historical, or natural or real world data is needed
major pro and con of archival data
you didn’t collect the data, so you doi’t have to spend time doing it but you also have little control over how it is done
secondary data
data from a preexisting study, but you are going to use it for something else
ethical considerations for archival data
IRB states that public info and materials are fair use, but it is important to consider that “participants” did not consent to being a part of the study
meta analysis
special kind of secondary data analysis that searches literature on a specific topic
Persuasion in a manuscript
You are communicating directly to the reader and should make it interesting
Shape of a manuscript
hourglass that starts broad, gets narrow, then broad again
abstract
what you are studying, hypothesis, results, and implications in very few words
Introduction
Bring in the reader with what the topic is, what has already been researched, and where the gap is
literature review
A logical progression of literature on the topic, showing what has already been studied, where the gap is, and what your hypothesis is.
What makes a good results section
a logical progression, including all relevant statistical information, reminding the reader of the hypothesis, and interpreting the analysis in lay terms
things that should go in tables
descriptive stats, t tests, correlation table, regression analysis
things that can be in figures
used for interaction results and mediation
formatting for figures and tables
item number, title, clarifying info, no color. can be included in the manuscript or at the end
Parts of the Discussion Section
hypothesis and findings, limitations, future research, implications of work
Reviewing hypothesis
clear explain the findings and relate to past research
limitations discussion
identify methodological flaws (confounds, validity) and concept flaws ( something that wasn’t considered or new research). there should be a sollution to all limitations
Future reserch
highlight future directions, can be related to limitations or just what the next steps are
Implications
how does the study inform the knowledge of this area
pros of presenting at a conference
talk about project and interests, looks good on a resume, see the newest research in the field, talk to grad advisors and other people in the field