The flashcards below were created by user
on FreezingBlue Flashcards.
Linking theory and methods important questions
What form of theory are we talking about? And Is data being collected to test a theory or build one?
Theories of the middle range
- Represent attempts to understand and explain a limited aspect of social life; operate in a limited domain and vary somewhat in their application.
- Ie: Structural functionalism, symbolic interactionsim, critical theory etc
Grand theories according to Merton
Offer few indications to researchers as to how they might guide or influence the collection of empirical evidence ie: they are hard to test; very abstract; therefore are of limited usefulness in social research.
What was Merton�s goal?
To seek to clarify what is meant by theory when social scientist write about the relationship between theory and research
Research that has no obvious connection with theory and is often prone to being dismissed
Represents the commonest view of the nature of the relationship between theory and social research
Is In opposite movement of deduction when the researcher infers the implications of his finding for the theory that promoted the whole exercise.
An epistemological issue concerns what question?
What is regarded as acceptable knowledge in a discipline and should the social world studied according to the same principles, procedures and ethos as the natural sciences.
An epistemological position that advocates the application of the methods of the natural sciences to the study of social reality and beyond
What does positivism entail?
What two features does realism share with positivism?
- A belief that the natural and the social sciences can and should apply the same kinds of approach to the collection of data and to explaination.
- A commitment to the view that there is an external reality to which scientist direct their attention
Through the use of appropriate methods, reality can be understood
A specific form of realism whose manifesto is to recognize the reality of the natural order and the events and discourses of the social world that holds that �we will only be able to understand- and so change- the social world if we identify the structure at work that generate those events and discourses.
A term given to a contrasting epistemology to positivism; subsumes the views of writers who have been critical of the application of the scientific model to the study of the social world and who have been influenced by different intellectual models.
-concerned with the empathic understanding of human action rather with the forces that are deemed to act on it.
-a philosophy that is concerned with the question of how individuals make sense of the world around them and how in particular the philosopher should bracket out preconceptions in his grasp of that world; a main intellectual tradition that has been responsible for the anti-positivist position
-an alternative to the positivist orthodox; a view that a strategy is required that respects the differences between people and the objects of the natural sciences and requires that scientist grasp the subjective meaning of social action. Hermeneutic phenomenological tradition->Symbolic interactionism
- -concerned with nature of social entities
- -question of whether social entities can and should be considered objective entities that have a reality external to social actors (objectivism) or whether they can and should be considered social constructions built up from the perceptions and actions of social actors (constructionism)
-an ontological position that implies that social phenomena confront us as external facts that our beyond our reach or influence
-an alternative to objectivism; an ontological position that asserts that social phenomena and their meanings are continually being accomplished by social actors. Social phenomena and their meanings are in constant revision. The researcher always presents a specific version of social reality rather than one that can be definitive.
Quantitative vs qualitative research regarding theory
- Quantitative: theory precedes research
- Qualitative: research precedes theory
Three of the most prominent criteria for evaluation of social research
-Reliability, Replication and validity
Are the results of a study repeatable?
Quite rare in social research; however, replication is highly regarded, and proves reliability.
Concerned with the integrity of the conclusions that are generated from a piece of research.
Aka construct validity; whether a measure that is devised of a concept really does reflect the concept that it is suppose to be denoting.
- Concerned with the question of whether a conclusion that incorporates a causal relationship between two or more variables holds water. Can we be sure that x causes Y and not some other variable? Independent variable= the factor that has the causal impact
- Dependant= effect variable
Can the results be generalized beyond the specific research context? Can the results be applied beyond the research sample?
Can the finding be applicable to peoples everyday, natural social settings?
Viewing all objects of study, whether natural or social, as belonging to the same realm and a consequent commintment to the principles of the scientific method.; being true to the nature of the phenomena being studied; a style of research that seeks to minimize the intrusion of artificial methods of data collection.
Five types of design
Experimental, cross-sectional, longitudinal, case study and comparative
Rare in sociology but are the �gold standard�; includes manipulation and treatment and control; Eliminates threats to internal validity in insurance level of health example
A control group is compared to two treatment groups but no random assignment
Entails the collection of data on more than one case at a single point in time in order to collect a body of quantitative or qualitative data in connection with two or more variables, which are then examined to detect patterns of association; Ambiguity about causation/ direction what causes what.
Timely and costed therefore relatively little use; a sample is surveyed and then survey at least one other time on another occasion.
Two types of longitudinal design
Cohort and panel
Panel longitudinal design
Repeated observations on a sample drawn from the entire population
Repeated observations on members of one cohort (e.g., British Cohort Child Development Studies)
Advantages of Longitudinal Designs
- Partially distinguished between age, period, and cohort effects
- Examines individual change
Disadvantages of Longitudinal Designs
- Several years until data are completed
- In long-term studies, early measure may not meet current measurement standards
Case study design
The detailed and intensive analysis of a single case; Provide rich accounts of how communities, organizations, and groups work;
How are case studies useful/ how are they limited?
- Very useful to generate Hypotheses and An understanding of �how things work�
- Difficult to assess reliability and validity
When is internal validity threatened?
In non-experimental research
A replicable, scientific, and transparent process that aims to minimize bias through exhaustive literature searches of published and nonpublished studies by providing an audit trail of the reviewers decisions, procedures, and conclusions. Used more to study benefits rather than haphazard. attempts to provide same rigor as is found in research procedures
A systematic review that includes only quantitative studies. Includes summarizing the results of a large number of quantitative studies and conducting analytical test to show whether or not a particular variable has an effect.
Standard lit review
No rules for material selection, material may be omitted or disregarded.
Components of systematic review
Explicit, replicable procedures; Comprehensive and unbiased in coverage; Leaves an �audit trail�
Issues in Meta-Analysis
- Requires full reporting of variables that the meta-analyst would like to use
- Meta-analysis is not always feasible
- Easier to find published research than unpublished
- There may be biases in what research is included
- Systematic reviews and meta-analyses are usually an end in themselves
- Harvard approach to referencing is standard in sociology
- Avoid footnotes for references or for substantive material
- Passing off others' ideas or words as our own
- Includes paraphrasing
- Like passing counterfeit money
- It may be accidential, but one has a lot of explaining to do
The process of Quantitative research
- Theory->hypothesis->Research design->devise measurement of concepts-> select research sites
- ->select research subjects/respondants->administer research instruments/collect data-> process data-> Analyze data-> Findings/conclusions-> write up findings/conclusions
Used to tap concepts that are less directly quantifiable; something that is devised or already exist and that is employed as though it were a measure of a concept. Ie Job satisfaction
Advantages of multiple indicator measures
- It is possible that a single indicator will incorrectly classify many individuals.
- One indicator may capture only a portion of the underlying concept or be too general. Ie: job satisfaction according to pay, yet there is more involved in job satisfaction than just pay.
- You can make much finer distinctions
- How we actually measure a concept
- � E.g., education or some combination of income
- and education
- Another name for the measure
- � Most common: One indicator per concept
- � Better: Multiple indicators
- � Bryman uses �measure� for an indicator that is �unambiguous�
- (e.g., income) but that oversimplifies the measure of income (e.g.,
- forgotten items of income
Why Study Something We Already Know?
- � More precise statement of relationship
- � Better measures
- � Extend what is already known
- � E.g., Compare siblings to eliminate family effect
- � Examine change over time
- � Likert scale is typically used as additive
- � Better if each measure is reliable
- � Reliability means little random error
- � Throws out random error; keeps only �true score�
- � Draws out different dimensions
5 Ways to Evaluate Measurement Validity
- 1. Face validity
- 2. Concurrent validity
- 3. Predictive validity
- 4. Construct validity
- 5. Convergent validity
Although each of 5 ways to evaluate measurement validity seems reasonable what are the associated problems?
- � Each assumes we know correct relationship
- � In real world, nothing is perfectly correlated
The measure apparently reflects the content of the concept in question
The researcher employs a criterion on which cases are known to differ and that is relevant to the concept in question
The researcher uses a future criterion measure rather than a contemporary one.
The researcher is encouraged to deduce a hypothesis from a theory that is relevant to the concept
The validity of a measure can be gauged by comparison to measures of the same concept through other methods
Why does Validity assumes reliability and Unreliable measure cannot be valid?
- � Measure = true score + random error
- � A true score is reliable (by definition)
- � Random error is not reliable (by definition)
- � True score portion may be valid
- � If random error predominates, changes not related
- to anything
Preoccupations of Quantitative Researchers
- � Measurement
- � Causality - �explaining things�
- � Generalization
- � Replication
- � We infer causation; we do not observe it
- � Similar issue to �critical realism� in Chapter 1
- � Empirical (�na�ve�)realism avoids causation
Explain how Empirical (�na�ve�)realism avoids causation
- � We only know what we observe
- � Humans always search for causes to make the
- work explicable
- - Causal thinking is a product of the human mind,
- not external reality
- � �Causes� are unknowable
- � Hence we should only talk about association
J.S. Mills four methods to infer causation
- � Agreement
- - X always present when Y occurs
- � Difference
- - X not present when Y does not occur
- � Joint
- - First two methods together
- � Concomitant variation
- - Amount of X covaries with amount of
More modern approach to mills work:
- � When X present, higher probability of Y
- � Probabilistic interpretation
- � Allows for multiple causes
- � E.g., Y occurs when there are enough causes
- present to reach a threshold