Long history in the social sciences

Increasingly becoming an essential component in HSR

Enable us to access areas not amenable to quantitative research such as lay and professional health beliefs

Prerequisite of good quantitative research, particularly in areas that have received little previous investigation

Goal is to develop concepts which help us to understand social phenomena in natural (rather than experimental) settings

They do not seek to provide quantified answers to research questions

e.g. “What is X and how does X vary in different circumstances, and why? (rather than how many Xs are there?)

Randomization is the process of making something random; this means:

Generating a random permutation of a sequence (such as when shuffling cards).

Selecting a random sample of a population (important in statistical sampling).

Generating random numbers: see Random number generation.

Transforming a data stream using a scrambler in telecommunications.

Randomization is used extensively in the field of gambling (or generally being random). Imperfect randomization may allow a skilled gambler to have an advantage, so much research has been devoted to effective randomization. A classic example of randomization is shuffling playing cards.

Randomization is a core principle in the statistical theory of design of experiments. Its use was extensively promoted by R.A. Fisher in his book Statistical Methods for Research Workers. Randomization involves randomly allocating the experimental units across the treatment groups. Thus, if the experiment compares a new drug against a standard drug used as a control, the patients should be allocated to new drug or control by a random process.

Although historically "manual" randomization techniques (such as shuffling cards, drawing pieces of paper from a bag, spinning a roulette wheel) were common, nowadays automated techniques are mostly used. As both selecting random samples and random permutations can be reduced to simply selecting random numbers, random number generation methods are now most commonly used, both hardware random number generators and pseudo-random number generators.

Stratification is the building up of layers, and can have several meanings

Social stratification, is the dividing of a society into levels based on wealth or power.

Stratification in archaeology is the formation of layers (strata) in which objects are found.

Stratification of rock layers (strata) is part of the geologic field of Stratigraphy.

Stratification (botany), where seeds are pretreated to simulate winter conditions so that germination may occur.

In logic, stratification is a layering of predicate symbols to guarantee unique interpretations and to avoid paradoxical definitions like Russell's paradox.

In mathematics, Stratification has a separate meaning as applied to manifolds, and singularity theory, of a decomposition into pieces with specified relationships on fitting together.

In music stratification is a layering of musical texture or the independent operating of more than one parameter simultaneously (see auditory stream).

In meteorology, atmospheric stratification is the division of the atmosphere into distinct layers, each with specific properties such as temperature or humidity.

In histology, stratified epithelium refers to epithelium that consists of two or more layers of epithelial cells, in contrast to simple epithelium, which only has one layer.

It is an internationally recognised structure which enables those working to improve the health of local communities and apply Health for All principles, to meet and share information, research and experiences.

The Network was established in 1987. It currently receives a grant from the Department of Health (England).

In 1995 the Network became a Company Limited by Guarantee in order to affiliate formally to the World Health Organisation and other networks. In 1997 the Network became a registered UK charity.

Randomization is the process of making something random; this means:

Generating a random permutation of a sequence (such as when shuffling cards).

Selecting a random sample of a population (important in statistical sampling).

Generating random numbers: see Random number generation.

Transforming a data stream using a scrambler in telecommunications.

Randomization is used extensively in the field of gambling (or generally being random). Imperfect randomization may allow a skilled gambler to have an advantage, so much research has been devoted to effective randomization. A classic example of randomization is shuffling playing cards.

The Network was established as part of the World Health Organisation's "Health for All 2000" initiative, which was launched in 1981 in Europe. This strategy was based on three key principles of

Equity Community Participation Intersectoral Collaboration and more recently, Sustainable Development.

Action in Europe began in 1987 with the launch of the Healthy Cities Network. In 1998, Health for All 2000 was revised and relaunched as Health 21, providing a strategy for the 21st century.

The Health for All Network is a structure through which staff and members of intersectoral partnerships, individual agencies and individuals working to reduce inequalities in health can:

Exchange information and ideas Identify good practice Comment collectively on policy issues Contribute to local, national and intersectoral

debates on health improvement Promote their local work nationally Market locally developed resources

Secondly, it is a supportive mechanism for members which:

Co-ordinates a national database Publishes a quarterly newsletter Holds an annual conference Supports regional meetings Encourages and supports joint training Publishes resources Provides speakers and workshops at national

conferences.

"Health for All in Action" provides a review of activity in the UK. It describes the different structures that have formed since 1987 to tackle the overriding target of the WHO Health for All Strategy (Ottawa Charter, 1986), recently revised as Health 21, ( WHO, 1998) to reduce inequalities in health.

This review underlines the fact that there is no single way to implement Health for All at a local level. Just as communities are diverse, so too are local Health for All initiatives. Projects, initiatives and alliances in the UK have, therefore, followed different paths to achieve their objectives; and adopted different strategies and organisational structures.

Qualitative and quantitative approaches to research are intended to be different

It depends on the type of question that you’re asking (what sort of information you’re trying to find out)

e.g. RCTs are ‘gold standard’ for assessing effectiveness, but can’t be used for understanding the language people use or the way they think about the world

Quantitative research requires a large number of participants in order to achieve statistical power

It looks for a very narrow range of information from the large sample population

Qualitative research looks at smaller samples in more depth

If all the factors in a situation are like a threads in a weaving...

then, quantitative approaches measure the thread

that they think is the most significant, to find out about it’s effect

(proving or disproving a hypothesis)

and qualitative research looks a section of the weaving to understand what the threads are, how they all fit together and reflect the pattern overall

(hypothesis generating)

Some of the main qualitative methods currently used in HSR are:

Interviews (in-depth and semi-structured)

focus groups or ‘group interviews’

observation

case studies

Quantitative methods (including RCTs) have contributed to advances in the treatment of diabetes

However, health professionals may need answers to additional questions, e.g. those concerned with patient behaviour

This is where qualitative research can be useful

Can be conducted as an essential preliminary to quantitative research

Can be used to supplement quantitative work

Can be used to explore complex phenomena or areas not amenable to quantitative research

Two stage investigation of the geographical variation in the incidence of operations on the tonsils and adenoids (Bloor et al 1976)

Part one: Epidemiological study – documenting variations

- Analysis of 12 mths routine data on referral, acceptance, and operation rates for new patients

Part two: Sociological study – explaining how and why variations come about- observation of assessment routines undertaken in outpatient departments

Part one: Epidemiological study – documenting variations

Found significant differences between similar areas with regions in referral, acceptance, and operation rates that were not explained by disease incidence

Operation rates influenced, in order of importance, by:

- Differences between specialists in tendency to list for operations

- Differences between GPs in tendency to refer

- Differences between areas in symptomatic mix of referrals

Part two: Sociological study – explaining how and why variations come about

Found considerable variation between specialists in their assessment practices:

- “High operators” – tendency to view a broad spectrum of clinical signs as important / to assert the importance of examination findings over child’s history

- “Low operators” – tendency to give examination less weight in deciding on disposal /to judge a narrower range of clinical features as indicating need to operate

Can be conducted as an essential preliminary to quantitative research

Can be used to supplement quantitative work

Can be used to explore complex phenomena or areas not amenable to quantitative research

Much qualitative research is interview based

Types of interviews

- semi-structured (open ended questions)

- in-depth (or unstructured; one or two issues covered in great detail; questions are based on what interviewee says)

These can be contrasted with structured interviews

To discover the interviewee’s own framework of meanings

To avoid imposing the researcher’s structures and assumptions (as far as possible)

To remain open to possibility that emerging issues may be different from those predicted as outset

Qualitative interview studies address different questions from those addressed by quantitative research

A quantitative epidemiological approach to SIDS might measure statistical correlates of national and regional variations in incidence.

In a qualitative study you might interview mothers of young babies in different ethnic groups to understand their child rearing practices

Ref: Gantley M, Davis DP, Murcott A. Sudden infant death syndrome: links with infant care practices. BMJ 1993; 306:16-20

A quantitative study of singlehanded GPs might compare their prescribing and referral rates, out of hours payments, list sizes, and immunisation rates with those of GPs in partnerships

In a qualitative study you might examine the concerns of singlehanded GPs (semi-structured interviews) and identify problems perceived by this group of doctors

Ref: Green JM. The views of singlehanded GPs: a qualitative study. BMJ 1993; 307: 607-10

Qualitative interviewers try to:

- be interactive and sensitive to language and concepts used by interviewee

- go below the surface of the topic being discussed, explore what people say in as much detail as possible

- uncover new areas or ideas that were not anticipated at the outset of research

Behaviour or experience

Opinion or belief

Feelings

Knowledge

Background or demographic

The less structured the interview, the less questions are determined and standardised before the interview occurs:- list of core questions that define the areas to be covered - the order in which questions are asked will vary- as will the questions designed to probe interviewee's meanings

The opportunity to capture in depth the views of one person, letting them take the lead

Accessing language and meaning that might be missed if using a structured interview

Privacy of interview makes them suitable for generating data on sensitive topics

Interview format familiar to most in UK from popular culture

Each interview represents the views of only one person

One-to-one interviews might increase the possibility of acquiescence bias (interviewee saying what they think interviewer want to hear rather than what they really think or do)

Might be culturally inappropriate (to ask direct questions or meet one-to-one)

A form of group interview that capitalises on communication between research participants in order to generate data

Not just a quick and convenient way to collect data from several people simultaneously

Explicitly uses group interaction as part of the method

People are encouraged to talk to one another: asking questions, exchanging anecdotes and commenting on each others' experiences and points of view

Useful for exploring people's knowledge and experiences

And can be used to examine not only what people think but how they think and why they think that way

Popular method for assessing health education messages

Examining public understandings of illness and of health behaviours

Examining people's experiences of diseases and of health services

Effective technique for exploring the attitudes and needs of staff

Group processes can help people to explore and clarify their views in ways that would be less easily accessible in a one-to-one interview

Helps researchers tap into the many different forms of communication that people use in day to day interaction (jokes, anecdotes, teasing, arguing)

Can encourage participation from those who are reluctant to be interviewed on their own (those intimidated by formality and isolation of a one-to-one interview)

Can actively facilitate the discussion of taboo topics because the less inhibited members of the group break the ice for shyer participants

Possibility of ‘group-think’: when people all agree in public, even though they might think differently

Some people may be shy of speaking in public Some people may dominate the group

discussions Requires good facilitation skills and choice of

materials (resource intensive)

Flexible and powerful tools which can open up many new areas for research

Can enable researchers to investigate research questions which would be difficult to investigate with more quantitative methods

The status of all forms of research depends on the quality of the methods used

Considerable debate over whether qualitative and quantitative methods can and should be assessed according to same quality criteria

Can be assessed with the same broad concepts of validity and relevance used for quantitative research, but these need to be operationalised differently

Triangulation

Clear exposition of methods of data collection and analysis

Reflexivity

Compares the results from either two or more different methods of data collection (e.g. interviews and observation) or, more simply, two or more data sources (e.g. interviews with members of different interest groups)

The researcher looks for patterns of convergence to develop or corroborate an overall interpretation

A clear account of the process of data collection and analysis is important

Written account should include sufficient data to allow the reader to judge whether the interpretation offered is adequately supported by the data

Sensitivity to the ways in which the researcher and the research process have shaped the collected data

The effects of personal characteristics such as age, sex, social class, and professional status needs to be discussed

Attention to negative cases

Research can be relevant when it adds to knowledge or increases the confidence with which existing knowledge is regarded

Another important dimension is extent to which findings can be generalised beyond the setting in which they were generated- ensure research report is sufficiently detailed- theoretical sampling

Worth or relevance – was this work worth doing at all?

Clarity of research question – if not at the outset of the study, was the research question clear by the end of the research process?

Appropriateness of the design to the question?

Context – Is the context or setting adequately described?

Sampling – Did the sample include a wide range of possible cases or settings?

Data collection and analysis – were the data collection and analysis procedures systematic?

Reflexivity of the account – Did the researcher self consciously assess the likely impact of the methods used on the data provided?

Qualitative methods can, and do enrich our knowledge of health and health care

They are not superior to other methods, but we need a range of methods to hand if we are to understand the complexities of health care

A complex process that involves moving back and forth between concrete bits of data and abstract

concepts between inductive and deductive reasoning between description and interpretation

Simply put: Data analysis is the process of making meaning from the data

Explore the data by reading through all of your information to obtain a general sense of the information

Memo ideas while thinking about the organization of the data and considering whether more data are needed Jot memos in margins of fieldnotes, transcripts,

documents, photos

EDUC 7741/Paris/Terry

Coding data Developing a description from the data Defining themes from the data Connecting and interrelating themes

Open Coding Assign a code word or phrase that accurately

describes the meaning of the text segment Line-by-line coding is done first in theoretical

research More general coding involving larger segments of

text is adequate for practical research (action research)

The process of looking for categories that cut across all data sets

After this type of coding, you have identified your themes You can’t classify something as a theme unless it cuts

across the preponderance of the data

After open coding an entire text, make a list of all code words

Cluster together similar codes and look for redundant codes

Objective: reduce the long list of codes to a smaller, more manageable number (25 or 30)

EDUC 7741/Paris/Terry

Take this new list of codes and go back to the data Reduce this list to codes to get 5 to 7 themes or

descriptions Themes are similar codes aggregated together to

form a major idea in the database Identify the 5-7 themes by constantly comparing

the data (Constant Comparative Analysis)

A process whereby the data gradually evolve into a core of emerging theory

This core is a theoretical framework that further guides the collection of data

Major modifications are lessened as comparisons of the next incidents of a category to its properties are carried out (Merriam, 1998).

It is best to write a qualitative report providing detailed information about a few themes rather than general information about many themes

ThemesThemes can also be referred to as CategoriesCategories

The names can come from at least three sources: The researcher The participants The literature

Most common: when the researcher comes up with terms, concepts, and categories that reflect what he or she sees in the data

Reflect the purpose of the research Be exhaustive--you must place all data in a

category Be sensitizing--should be sensitive to what is

in the data i.e., “leadership” vs. “charismatic leadership”

Be conceptually congruent--the same level of abstraction should characterize all categories at the same level For instance, you wouldn’t have produce, canned

goods, and fruit

Ordinary: themes a researcher expects Unexpected: themes that are surprises and not

expected to surface Hard-to-classify: themes that contain ideas that do

not easily fit into one theme or that overlap with several themes

Major & minor themes: themes that represent the major ideas, or minor, secondary ideas in a database Minor themes fit under major themes in the write up

A detailed rendering of people, places, or events in a setting in qualitative research

Codes such as “seating arrangements,” “teaching approach,” or “physical layout of the room,” might all be used to describe a classroom where instruction takes place

From the coding and the themes, construct a narrative description and possibly a visual display of the findings for your research report

Use the assigned format (see syllabus)

Identify dialogue that provides support for themes

Look for dialogue in the participants’ own dialect Use metaphors and analogies Collect quotes from interview data or

observations Locate multiple perspectives & contrary evidence Look for vivid detail Identify tensions and contradictions in individual

experiences

Because qualitative researchers believe that personal views can never be kept separate from interpretations, personal reflections about the meaning of the data are included in the research study “David had been diagnosed with AD/HD and also with mild Tourette

Syndrome. He took medication for AD/HD. He was selected to participate in the project as a confirming participant because he was so involved with the project and so intense during the first observation. Unaware that he had AD/HD and Tourette Syndrome until I interviewed his mother during the second year of the project, I was surprised because he was the most focused student in the classroom.”(Terry, 2003)

Qualitative researchers often display their findings visually Comparison table or matrix Hierarchical tree diagram that represents themes

and their connections Boxes that show connections between themes Physical layout of the setting Personal or demographic information for each

person or site

Enhances Commitment, Attitudes,

and Student Development

Interpret the data in view of past research Show how the findings both support and

contradict prior studies “These findings are consistent with other studies in regard to duration.

It has been found that the length or duration of service learning projects has an impact on student outcomes, with the longer duration projects having greater impacts. However, significant differences are not found in projects lasting over 18 weeks (Conrad & Hedin, 1981). The project on which this study focused was examined over a year and a half period of time; thus it is considered to be long in duration which helps to explain its impact on student outcomes.”

The researcher suggests possible limitations or weaknesses of the study “This study focused on one rural middle school in an area in

Northeast Georgia, Hartwell. It documented the methodology used in the service learning project and the effect of a certain type of service learning model, Community Action. Therefore, the study provides an in-depth look at a service learning project carried out by gifted students in just one middle school in a rural area situated in a Southern state. Transferability may be limited as a result” (Terry, 2001).

Researchers make recommendations for future research “In addition, further research is needed to determine outcomes for

a diversified culture of students, including, but not limited to African-American students and students diagnosed with AD/HD. Research is also needed to examine and validate existing frameworks before professing any general claims concerning the outcomes for students engaged in service learning activities” (Terry, 2003).

At the end, the qualitative researcher validates the finding by determining the accuracy or credibility of the findings. Methods include: Prolonged engagement & persistent observation in the

field Triangulation Peer Review Clarifying researcher bias Member Checking Rich, thick description External Audit

“I am not an impartial bystander when it comes to service learning so I knew I had to enhance internal validity at the outset of the study. I have been involved with Community Action service learning projects for over 16 years as a teacher. I have co-authored a book on how to facilitate Community Action service learning projects which I have used to implement service learning projects in my own classroom. My students have been featured in Reader’s Digest and have been guests on the Phil Donahue Show because of their outstanding work in service learning. Being aware of this bias, I took extreme precautions to maintain objectivity during both the collection and analysis of the data thereby accurately representing the project fairly and accurately” (Terry, 2001).

From a quantitative perspective, reliability refers to the extent to which research findings can be replicated

From a qualitative perspective, dependability, (reliability) in qualitative research is not based on outsiders getting the same results, but that outsiders concur that, given the data collected, the results make sense. In other words, the results are dependable and consistent (Lincoln & Guba, 1985).

Concerned with the extent to which the findings of one study can be applied to other situations

Quantitative studies enhance external validity using a priori conditions which are limiting in conducting qualitative research

External validity is problematic in qualitative research because “In qualitative research, a single case or small nonrandom sample is selected precisely because the researcher wishes to understand the particular in depth, not to find out what is generally true of the many” (Merriam, 1998, p. 208).

Think in terms of the reader of the study What is the extent to which a study’s findings can

apply to other situations? This is referred to as Representativeness or Representativeness or

TransferabilityTransferability Merriam (1998) suggests: rich, thick description and

typicality, modal category, or multisite designs“To enhance external validity in this study, these procedures were followed: