Midterm Exam Flashcards
1
Q
What is social science research?
A
- The social component refers to human behaviour
- The science component refers to the methodological approach by which
human behaviour is studied - Meaning social science refers to the study of human behavior following
some set of methodological principles
2
Q
Epistemology
A
the study of the nature, scope and production of human knowledge. It is concerned with questions like “what is knowledge?”
3
Q
positive epistemology
A
maintains that an object is reality in the world independant of the observer, that is that reality can be observed
- this view allows the study of social science
4
Q
Non- positivist epistemology
A
maintains that reality does not exist independent of human perception and that it is impossible to study phenomenon without influcing it.
5
Q
the scientific method
A
a set of procedures used to test hypotheses about phenomena based on the collection and analysis of data through observation and interaction
6
Q
6 Steps of the scientific method
A
- problem identification
- hypotheses formulation
- measurement
- data collection
- data analysis
- confirm or disconfirm analysis
7
Q
Violations in the scientific method
A
violations in the method arise any time a subsequent step occurs prior to an antecedent step.
e.g letting the decide what hypotheses to use
8
Q
Research ethics
A
the norms standards and legal rules regarding the appropriate behavior in the conduct and publication of research
9
Q
what constitutes a good research topic
A
a good research topic should be important and interesting to the researcher, in other to sustain interest during the research process, and in order to interest others
10
Q
A research question is important when…
A
- it challenges conventional wisdom
- there is a significant amount of news, interest and academic research on the issue
- it has an impact on a large amount of people
- it has concrete implications on how to better the world
- when it has important applications to other issues than the one explored.
11
Q
Research question vs topic
A
- question is much more specific than a topic
- question relates to a particular aspect of a broader topic that a researcher seeks to understand
12
Q
What is a research puzzle?
A
a puzzle is a question for which the answer is not immediately obvious.
13
Q
theoretical contribution
A
relates to the explanation provided for a given
question or puzzle
- Refine an existing theory
- Refute or argue against an existing theory
- Develop a new theory to explain a phenomenon.
14
Q
empirical contribution
A
refers to the evidence offered in support of this
explanation
- test an un-tested hypothesis
- confirm or challenge an existing theory using different data and/or methodological techniques
15
Q
4 different approaches to research
A
- problem driven
- funding driven
- data driven
- methods driven
16
Q
purpose of a literature review
A
summarize and synthesize the existing academic (and non-academic) literature on a particular subject in order to characterize the strengths and weaknesses of the literature
17
Q
what is a concept
A
most basic formation - abstract idea or object (e.g democracy, war etc)
18
Q
why do concepts matter?
A
- concepts establish the defining features or properties of political, social, psycological and economic phenomena. They provide the basis around which arguments related to these phenomena are constructed and have significant implications, as a result for how the world is understood
- how a phenomenon is conceptualized is important because defined in 1 way, a concept may be unrelated to another phenomenon, but defined in a diffrent way, it may be strongly related to it
19
Q
characteristics of good concepts - 5 criteria to evaluate concepts
A
- clarity
- delineation
- scope
- coherence
- discriminatory power
20
Q
6 step procedure to concept building
A
- think abstractly about what the object that you want to conceptualize is without reference to the ideas of others
- examine real world examples of the object
- look closely at prominent definitions in your discipline of the object you are conceptualizing
- once again, examine real world examples of the object that you are studying
- evaluate quality of your concept against the criteria of good concepts
- justify why your concept is diffrent from and preferable to other prominent defentions of the concept
21
Q
concept stretching
A
- distortion of a concept due to its application to other causes
- concept is disorted to point it looses its original meaning
- when concepts are distorted they can be broadened to include a case which they should not apply or narrowed to discluded a case which they should apply
22
Q
deductive reasoning
A
- top down approach in which researchers derive specific, testable hypotheses to explain human behaviour from theoretical axims
- important advantage - facilitates cumulative knowledge building
- researchers using it base it base their work on existing theories by; filling in gaps and holes in existing theories, identfying new condtions in which existing theories apply, using the basic logic of existing theories to explain phenomena yet to be explored
23
Q
Inductive reasoning
A
- bottom up approach in which researchers conduct generalizable theories from first observing patterns in human behaviour and then developing hypotheses to explain these patterns
- important advantage - researchers arguments are grounded in actual experiences so that the assumptions and theories they derive from it are realistic
- potential disadvantage - arguments derived may be more descriptive of the cases from which they are derived then theoretical
- also run the risk that arguments derived are specfic to the case they are derived from
24
Q
Sufficient condition
A
- an explanatory factor that alone is enough to produce a given outcome
- not also nessessary condtions because outcomes can be procuced in the abcense of certain sufficent condtions
25
necessary condition
- an explanatory factor that is needed to produce a given outcome
- diffrent because it may not be enough alone
26
relationship directions
- the direction of the effect refers to; weather an explanatory factor is likely to increase or decrease the outcome and weather this effect is linear, non linear, monotonic or non monotonic
27
tautology
- a statement that is true by definition
- convey no information because the antecedent is the same as the postcedent, they also violate falsifiability
- reverse causation refers to the situation where the explanatory variable does not cause the outcome variable, but the outcome variable causes the explanatory variable
28
Method Selection
researchers have to choose between qualitative and quantitative approaches or mixed method researches (combines the two)
29
Qualitative research
- involves the collection and analysis of data based on written, verbal and visual content
- encompasses a number of different methods, including interviews, focus groups, participant observation, an process tracing
30
Quantitative research
- involves the collection and analysis of numeric data, some of which may be derived from assigning numerical values to qualitative info
- experiments, surveys and observational studies
31
Mixed methods research
involves collecting, analysing and interpreting quantitative and qualitative data in a single study or series of studies examining the same phenomenon
- overall objective is too combine the strengths of both approaches to build more accurate and precise theories, conduct deeper and more rich analyses and to produce greater confidence in the results of studies
32
disadvantages and advantages of different methods of research arise around 5 key aspects of research
- hypothesis building
- hypotheses testing
- causal inference
- generalizability
- replicability
33
forms of mixed methods research design
- design
- Concatenation
- Gapʿfilling
- Triangulation
- Interpretation
34
case selection
- prior to collecting data, researchers must determine how many and which cases to include in their analyses
- a case is an instance of a class of events
35
2-4 steps for the process of case selection
1. determining the number of cases to be included in the analyses based on the goals of the study
2. selecting the cases through either random or non-random sampling techniques
and for cases involving human subjects
3. recruitment of subjects
4. retention of the subjects
36
How many cases to study
- a researcher must choose, a single case, mid range number of cases, or a large number of cases
37
single case studies
- excellent option when the goal of the researcher is to develop a new theory to explain an existing problem/phenomenon
- allow researchers to dig deeply into the details of a case and develop a rich compelling argument
- disadvantage - theory developed through a single case has a lower likelihood of being generalizable and cannot be used to confirm or disconfirm probabilistic theories
- much less valuable for theory testing then theory building
38
midrange case studies
- include more than one case but fewer than are needed to detect statistically significant effects (there is no absolute #)
- more generalizable and allow for broader theoretical development than single case studies
39
large-N case studies
- defined as any study that includes the number of cases nesseassary to dectect statistcally signifcant effects if one exists
- are appropriate for theory testing
- can be used to confirm or disconfirm probabilistic theories
- analysed using statistics
40
case selection methods
- Researchers resort to a subset of cases or observations of a target
population of interest to test for their hypotheses.
- The target population is the group of individuals or cases from which
researchers draw their samples.
- To identify this subset of cases (sample), researchers may use either
non-random or random sampling techniques.
41
Random sampling techniques
➔ Random sampling is the gold standard for data collection. It is necessary,
but not sucient, for constructing representative samples.
➔ In random sampling, every person or case has a known probability of
being included in the sample.
42
random sampling techniques: Simple random sampling
every member (or case) in the target
population has a known, equal, non-zero chance of being selected
43
random sampling techniques: Systematic random sampling
every kth element (i.e., ratio of the sample
size to the target population) of the target population is sampled. The
index of the starting element is then selected within the first k elements by
SRS. Multiple of k are added to the starting index to form indices for
subsequent elements
44
random sampling techniques: Stratified sampling
simple random samples are taken in proportion to the
population from subgroups (e.g., regions, genders, religions, etc.) of the
target population
45
random sampling techniques: Clustered sampling
simple random samples are selected from subgroups
of the target population. Samples are only taken from certain subgroups
of the target population that are determined by SRS
46
random sampling techniques: Over/undersampling
selection of more/fewer individuals from a subpopulation of the target population known to participate at a lower/higher
rate to ensure that the sample is representative of the target population
47
Non-random sampling techniques
➔ Non-random sampling techniques are a set of techniques in which each
case or observation in the target population does not have a known, nonzero chance of being included in the analysis.
➔ Samples constructed through non-random sampling techniques are less
likely to be representative of the target population than random sampling
techniques
➔ Researchers may use non-random sampling techniques when:
➔ their goal is to gather information from a range of cases in order to
develop hypotheses about human behavior;
➔ they do not have the funds or resources to use random sampling methods;
➔ they are unable to identify a sampling frame, namely the target population
from which to draw a random sample of participants.
48
non-random sampling techniques: Convenience
respondents selected based on ease of accessibility
49
non-random sampling techniques: Quota
respondents selected in proportion to the target population
with respect to known characteristics
50
non random sampling techniques: snowball
respondents selected based on recommendations of prior
respondents
51
non random sampling techniques: Volunteer
respondents self-selected into survey
52
Selection bias
➔ Selection bias occurs when observations are chosen according to some
rule that is correlated with the dependent variable.
➔ Selection bias may be a result of an aspect of the method by which
researchers select their cases, or it may be a function of nature. It is more
common of non-random sampling techniques.
➔ Selection bias is problematic for research because it results in a sample in
which there are a disproportionate number of cases that are more or less
likely to experience the outcome.
53
