Research Methods and Stats Flashcards
A great deal of power that comes from the use of statistical knowledge to prove a point comes from
The fact that much of the public is under-educated on what these figures and ratios mean.
Statistics can act as a
subjective mechanism of social control when decisions of researchers are made in error
Conceptualization
identifying what we mean when using specific terms/concepts
Standardization of data
Presenting data in a way that allows comparison of different measures
Controversies over Counting Crime
Criminologists have developed theories and conducted research using crime statistics that may not be entirely valid or reliable.
Statistics and theories of crime can be misused. In sum, we need to critically analyze crime statistics for their validity and reliability.
Surveys
series of closed-ended questions designed to focus the respondents’ answers in a comparable manner
A deductive approach
- Instrument development e.g. questionnaire, interview schedules
- May be conducted in person, via the telephone, via mail
Every time a survey is completed, the information is contributed toward:
datasets used by stakeholders to produce knowledge about the public
Likert-style scaling
reading a proposition paired with a ranking system that uses a scale of some sort
Good survey questions avoid:
- Jargon or abbreviations
- Ambiguity
- Emotional language
- Double-barreled questions
- Leading questions
- Questions that are beyond respondents’ capabilities
- False premises
- Asking about future intentions
- Double negatives
Validity
Ability of research tools to effectively represent the attitudes and ideas that they are meant to measure
8 separate tests of validity:
- Face
- Content
- Criterion
- Concurrent
- Predictive
- Construct
- Convergent
- Discriminant
Inferential limitation
The extent to which numerically derived statistics can be used to make inferences on a broader population
Construct Validity
A measure of how well a series of questions can be used to make inferences about a given topic
- Jeopardized when researchers either use ineffective measures for their identified trait or develop constructs that identify poorly defined traits
Content Validity
A measure of how well a series of questions represents the concept they are designed to test
- Underlying theory is that good research replicates the successes of the past and discards the failures, creating an environment of continual improvement
- Above assumption becomes increasingly difficult to ascertain with deeper consideration of the meaning of expertise
Likert Scale
Asks respondents to report on the intensity of different items through the use of standard response categories (strongly agree, agree, disagree, strongly disagree)
- PRoduces responses that can lead to an average score on the concept
Sampling
- Process of selection from the population
- Samples represent the population
Sampling bias
A critique of statistical research based on how well a sample depicts the population it is meant to represent -> the methods to locate and contact would-be survey respondents are flawed based on the way the survey is administered.
Haphazard sampling
techniques that intentionally employ non-random sampling strategies
Tactics: convenience and purposive samples
Convenience Sampling
Collecting responses from a group of people who are readily available to the researcher (i.e public place as people walk by)
Purposive Sampling
Seeks out individuals who represent a specific group within the population
Non-probability sampling
- Haphazard/Convenience sampling
- Quota sampling
- Purposive sampling
- Snowball sampling
Non-probability: Haphazard/Convenience Sampling
- Selecting convenient cases
- Not random, may misrepresent the population
Non-probability: Quota Sampling
- Selecting a certain proportion of a defined group from the population
- Can ensure that certain characteristics are represented
Non-probability: Purposive Sampling
- Selecting cases based on the expert opinion of who needs to be selected
- Unique cases
- Specialized populations
- In-depth investigations
Non-probability: Snowball Sampling
- Selecting cases that are linked together
- Used to target interconnected networks
Big Data
Collection and analysis of personal information produced during our interactions with modern technology (automated, digital process)
Problems with Big Data
- Latent Bias
- Representativeness
Latent Bias
- Exclusion of marginalized populations
- Underlying computational models reflect the inherent bias of their creators, intentionally held or not
Representativeness
Digital invisibility: lack of access to internet and devices due to financial costs, affecting the accuracy of big data
Representing Unheard Voices
Statistical knowledge paired with opportunities to share the experiences of marginalized communities can be a valuable tool for overcoming systemic mechanisms of exclusion
Mixed-methods Research
- Approach to research that unites the strengths of qualitative and quantitative information gathering
Qualitative Data Analysis
“The non-numerical examination and interpretation of observations, for the purpose of discovering underlying meanings and patterns of relationships.”
- Ultimate goal is to identify patterns
- Involves some form of inductive reasoning
Finding patterns
- When we are looking for patterns we are concerned with similarities and differences in what we observe
- The similarities represent the patterns of interactions and events that are common to the phenomenon being studied
- The differences are important in understanding what characteristics are common among those who are different
Grounded Theory
An inductive approach to research in which theories are generated from an examination of data, through the constant comparing of unfolding observations
- Theories are comprised of well-defined categories that are developed based on the observations made about the relationships that exist as a part of some social phenomenon
- Conceptualization involves developing or refining concepts that are based on the data observed
Ethnography/Participant Observation
Refers to a report of social life that focuses on detail and rich idiographic account
- Data collected in the natural setting (naturalism)
- Involved naturalistic and holistic understandings of social reality
Case Study
A focused, details investigation of a single instance of some social phenomenon