research meth Flashcards
empirical approach
observe and measure- quantitative
interpretive approach
observe and interpret- qualititative
Critical thery
asking whose interests are advanced
psa
publis service advertisement
triangulation
using different methods to get a complete understanding
q methodology
used to systematically study ppls subjective viewpoints
World view 1
Nomothetic- communication as objectively predictable, measurable, generalizable
World view 2
Idiographic- communications as subjective, unpredictable, individual
Postpositive
the world is governed by laws and theories
Constructivist
individuals construct their own view point
Transformative
mixes research with politics
Pragmatism
focused on solutions
Open ended research questions
non directional- you want to know if there is a relationship
close ended
directional- you want to know what direction is the relationship
Two tailed hypotheses
non directional- you say there is a relationship but nothing about its type
one tailed hypotheses
directional- you say there is a relationship and what type it is, as A increases B decreases
Null (HO)
no relationship
Induction-
from observation to theory
Deduction
from theory to observation that will test the theory
Abduction
you start with the effect and you try to reason back to the possible cause
Ethical issues of research
Honesty, Confidentiality, Debriefing, (not) making generalizations, literature review, acknowledging others, appropriate language, no plagiarism
Judeo-Christian ethic
Would you participate in your own study?
Categorical impretive
a behavior is valid if it is seen as a universal rule
Principal of utilitarianism
research that hurts minority is justified if there is a greater good
Veil of ignorance
you act like you do not know not to cloud your judgement
Nuremberg code
benefits of the research must outweight the risks
Belmont report
emphasizes autonomy, beneficiency and justice
Basic assumption peer review
person evaluating your research are qualified enough to do a similar work as yours
Formal review
Official boards and organizations
Informal review
Networking…
NOIR
Nominal Ordinal Interval Ratio
Validity
does the measure capture what we are interested in?
Relability
can we be sure that if we do the same study again we will get the same results
Content validity
judgement call
content of face
i think it looks ok
Construct validity
do i think it is theoretically ok
Criterion validity
does it test ok
Test retest reliability
repeating the research to check if the results come out the same
intercoder/observer reliability
two observers of the same thing at the same time record the same results
Likert scale
from strongly disagree to strongly agree
Semantic differential scale
expresses opposite ideas towards the concept
Population
everybody in the group you are studying
Sample
part of population selected for the study
Census
study of the entire population
Sampling frame
list from which the sample is drawn
Sample units
units selected for the study eg couples, individuals, organisations
Piloting
testing the sample to identify possible problems
Non probability sampling
Convenience, purposive, quota, snowball, volunteer
simple random sampling
loterry
stratified random
making sure all the relevant groups are included in the sample
Multistage cluster
moving from bigger to smaller units
to show that A caused B we must show:
Time order, Meaningful co-variance, Non-spuriosness
Multivariate analyses
3 or more variables in total
Factorial design
more than 2 independent variables
Spurious relationships
relationship found between 2 variables but not the one you’ve been looking for
External validity
can we take the findings of the research outside of the study to the general population
Cross sectional survey
slice of time, you ask a person once and have no reason to believe they’ll change their mind over time
Longitudinal
surveys taken in several points over time
Dichotomous questions
select yer or no, can cause complications
Double barreled question
you ask two questions in one
leading questions
they lead the respondent to a certain answer
Framing
when you frame the question in a context that might affect the answer
Negative wording
people often miss the negation while reading the questions which can cause problems
Double negative
mix of double barreled and negative wording
Funnel structure
from general to detailed questions
inverted funned structure
from detailed to general questions
Branching questions
if your home has wifi go to question no. 6 if it doesn’t go to questions no. 10
Filter questions
they filter out the responses
observational studies
unobtrusive measures, observing ppl without them being aware of that
Performance studies
studying human behaviour in performance
Stemming (content analysis software)
changing variations of words to its stem form
Lemmatization (content analysis software)
grouping words together based on their basic dictionary definition
Unstructured interview
letting the interviewee tell the story, no fixed questions, like a conversation, useful to explore a new area
Structured interview
questions are prepared and fixed, no changes possible
semi structed interview
topics to cover are listed, there is room for a follow up question
