Week 1-6 (Qualitative) Flashcards
Ontology
Assumptions about the nature of reality
Epistemology
Set of assumptions about the relationship between the knower & known
Methodology
- Theoretically informed approach to how we gain knowledge about world
- Reflect rationale & philosophical assumptions underlying study
Induction
- Begins with concrete details
- Works toward abstract
Theory building
Deduction
- Begins with abstract ideas
- Works toward concrete -details
- Theory testing & refinement
Exploration
- Uncover phenomena of interest
- What kinds of things are present here
- How are things related
Description
- Describing phenomena of interest
- Precise & trustworthy
Comparison
- Shared features
- Not shared features
- Experiential vs consensual knowledges
Descriptive Variable
- Do not emphasize quantity of information
- Number may be used for data coding
Indirect Observation
- Studying traces of human behavior
- Analyzing archival data
- Secondary analysis
Disadvantages of Archival/Secondary Data
- Possible lack of authenticity
- Lack of representativeness
- Measurement error
- Recall bias
- Can’t confirm interpretations
Advantages of Archival/Secondary Data
- Reducing cost
- Avoids duplication of primary studies
- Elaboration of earlier findings
- Avoids overburdening informants
Direct Observation
- Watch people
- Record their behaviors
- Continuous monitoring
- Spot observation/time sampling
Elicitation
- Interviewing
- Structured/unstructured
Purposive Sampling
- Maximum variation
- Snowball - hard to find cases
- Stop when we hit theoretical saturation
Thematic Analysis
- Preliminary observations
- Identifying themes
- Developing a coding scheme
- Coding the data
Narrative Analysis
- Story telling - 1st person
- Signature story
- Rich data
Credibility
Credible/believable from the perspective of participant in research
Transferability
Generalized/transferred to other contexts/settings
Dependability
- Would the same results be obtained if we could observe the same thing twice
- Reliability
Confirmability
Results could be confirmed/corroborated by others
Selection Effects
Studied sample does not represent population of interest
Reactive Effects
Participant behaves different in the study then they would normally
Methodological Effects
Changes in results from different data collection procedures
Simultaneity
Analysis can begin once we start sorting out data
Metacoding
Sorting chunks of data according to theoretical concepts
Content Analysis
- Studies documents & communication artifacts
- Quantifies codes & themes
Content Analysis Key Concepts
- Systematic observation of data
- Codes assigned to indicate meaningful patterns
- Determine major common themes in content
- Show relationship between major themes
Discourse Analysis
- Studies language used to reveal values & beliefs
- Examines underlying significance of language within larger social context
Discourse Analysis Key Concepts
- A belief, practice, knowledge
- How language is used in a specific context
Rigour
- Validity of analyses & results in research
- No consensus on standards - difficult to demonstrate
- Needs justification to not be seen as personal opinion
Selection Bias
- Sampling people, time, place
- Choice of questions
- Data selection
Wording Bias
Encouraging particular response
Confirmation Bias
- Selecting participants data to confirm a belief
- Researcher’s personal bias
Halo Effect
Something is considered positive due to a single positive attribute
Triangulation
Learn more by observing multiple perspectives
Population
All members of a particular group of interest
Sample
Subset of population that is studied
Variable
Anything that has quantity/quality that varies
Independent Variable
Manipulated or chosen my researcher
Dependent Variable
Outcome variable
Control Variable
Held constant to test relationship between dep & indep
Nominal
Attributes are merely different from each other
Ordinal
Categories you can rank order along some dimension
Interval & Ratio
Attributes are rank ordered and have equal distances between adjacent attributes
Sensitivity
How effective is tool in detecting what you are looking for
High Sensitivity
- Low false negative rate
- Correctly identifies people with a condition
Specificity
How accurately does the tool give you a negative result when what you are looking for is absent
High Specificity
- Low false positive rate
- Correctly identifies people without a condition
Missing Completely at Random (MCAR)
- No pattern in the missing data on any variables
- Best case scenario
Missing at Random (MAR)
- There is a pattern in the missing data
- Not based on your primary dependent variable
Not Missing at Random (NMAR)
- Missing value depends on variable that is missing
- Worst case scenario
Validity
Are you measuring what you intended to measure
Reliability
Are you consistently getting the same response
Normality
- Main assumption of many statistical tests
- Check before testing research hypothesis
- Symmetric bell curved shape - over bar graph
Linearity
- Whether a relationship between 2 variables forms a straight line
- Can affect type of statistical tests used
Frequency Distribution
How many times each value appears for a given variable
Quartiles
Observations into 4 equal parts
Bar Charts
- Display nominal data on x-axis
- Spaces between bars to show distinct categories
Histograms
- Bar graph where vertical bars touch
- X-axis - continuous variables
- Y-axis frequencies/%
- Highest bar is the mode
Line Graphs
- Show how variable changes over time
- X-axis - time
- Y-axis - quantity of variable
Scatterplots
- Each dot represents indep & dep of a single case
- Number of dots is sample size
- Dots indicate direction & magnitude of relationship
