7 Data analysis 7 Data analysis Main aim

  • Slides: 39
Download presentation
7 Data analysis

7 Data analysis

7 Data analysis Main aim To present the data collected for a Focus Assessment

7 Data analysis Main aim To present the data collected for a Focus Assessment Study as a description of behaviour from the perspective of informants.

Data analysis To avoid bias and subjective interpretation of the data: only trained members

Data analysis To avoid bias and subjective interpretation of the data: only trained members of the research team should conduct the analysis two team members should separately conduct the analysis, compare their results, and discuss and find a solution to any differences.

Categorising and coding data • Prepare a template of responses, each firmly related to

Categorising and coding data • Prepare a template of responses, each firmly related to the research aim • Give each category a series of codes to denote aspects of the issue under investigation • Allocate each relevant section of the transcription to a code • Repeat for every category

Categorising and coding data Inductive category formation Examine data for themes that arise and

Categorising and coding data Inductive category formation Examine data for themes that arise and form categories from those themes. Deductive category application Formulate categories from expected responses.

Categorising and coding data • Informants do not talk in neat categories • Categories

Categorising and coding data • Informants do not talk in neat categories • Categories may need to be revised • ‘Unusable’ data

Content analysis of coded interview data Will reveal: • number of informants who thought

Content analysis of coded interview data Will reveal: • number of informants who thought that peer influence was a factor influencing drug use • number who did not • number who were unsure.

Content analysis of coded focus group data Will reveal: the relative importance of this

Content analysis of coded focus group data Will reveal: the relative importance of this theme compared to other themes: • how many comments • how many participants discussed it

Interpreting results To support recommendations for further research and/or interventions.

Interpreting results To support recommendations for further research and/or interventions.

Interpreting results Discourse analysis: interpreting what informants mean 'Nobody tells me what to do.

Interpreting results Discourse analysis: interpreting what informants mean 'Nobody tells me what to do. If I want to use drugs, I will - if I don't, I won't. As it happens, I do - because I want to' = high self-esteem? 'If your friend is smoking weed [marihuana], you wanna do it too because you don't want to feel left out. So you are gonna do it too' = low self-esteem?

Interpreting results DO NOT: • use discourse analysis unless trained • interpret results to

Interpreting results DO NOT: • use discourse analysis unless trained • interpret results to suit the views of the research team • interpret results to support the recommendations of the research team • ignore ‘inconvenient’ deviations from the norm

Interpreting results Ask for verification of interpretations and clarification of ambiguous or unexpected findings

Interpreting results Ask for verification of interpretations and clarification of ambiguous or unexpected findings

Analysis of other data • from observation

Analysis of other data • from observation

Analysis of other data • from observation • secondary data • supplementary questionnaires •

Analysis of other data • from observation • secondary data • supplementary questionnaires • researchers’ fieldnotes

Triangulation The use of multiple methods and an analysis of their findings in order

Triangulation The use of multiple methods and an analysis of their findings in order to yield a more comprehensive picture of the phenomenon under investigation.

8 Ethical issues • Informed consent aims of study how the study will be

8 Ethical issues • Informed consent aims of study how the study will be conducted benefits of participation how results reported how confidentiality maintained

Informed consent: an example From a questionnaire used by Fountain J, Howes S (2001):

Informed consent: an example From a questionnaire used by Fountain J, Howes S (2001): Rough sleeping, substance use and service provision in London. Final report to Crisis. London: National Addiction Centre 'This study is being conducted by Crisis [a wellknown organisation working with homeless people] and is about homelessness and drug use. I will be asking you and other homeless people questions about both these things, although you don't have to be using drugs to take part in the study. A report on the findings will be used to inform those providing services for homeless people who also use drugs.

Informed consent: an example From a questionnaire used by Fountain J, Howes S (2001):

Informed consent: an example From a questionnaire used by Fountain J, Howes S (2001): Rough sleeping, substance use and service provision in London. Final report to Crisis. London: National Addiction Centre 'This study is being conducted by Crisis [a well-known organisation working with homeless people] and is about homelessness and drug use. I will be asking you and other homeless people questions about both these things, although you don't have to be using drugs to take part in the study. A report on the findings will be used to inform those providing services for homeless people who also use drugs. I can assure you that everything you tell me is completely confidential: you need not give me your name and no information which can identify you will be passed on to anyone outside the research team.

Informed consent: an example From a questionnaire used by Fountain J, Howes S (2001):

Informed consent: an example From a questionnaire used by Fountain J, Howes S (2001): Rough sleeping, substance use and service provision in London. Final report to Crisis. London: National Addiction Centre 'This study is being conducted by Crisis [a well-known organisation working with homeless people] and is about homelessness and drug use. I will be asking you and other homeless people questions about both these things, although you don't have to be using drugs to take part in the study. A report on the findings will be used to inform those providing services for homeless people who also use drugs. I can assure you that everything you tell me is completely confidential: you need not give me your name and no information which can identify you will be passed on to anyone outside the research team. You don't have to answer any of the questions if you don't want to, but please answer where you can. The interview will take about 40 minutes to complete. Do you have any questions about the study or the interview? [Interviewer should answer all questions]. Do you agree to be interviewed? '

Informed consent • When observing people in a public place? • When interviewing adolescents?

Informed consent • When observing people in a public place? • When interviewing adolescents? • Parental permission? • When an individual is encouraged by others to participate?

Ethical issues • Informed consent • Confidentiality Must be guaranteed Particularly important for those

Ethical issues • Informed consent • Confidentiality Must be guaranteed Particularly important for those asked about drug use / criminal activities

Maintaining confidentiality • When names / locations of drug dealers are revealed • When

Maintaining confidentiality • When names / locations of drug dealers are revealed • When storing data • In focus groups

Ethical issues • Informed consent • Confidentiality • Anonymity It must not be possible

Ethical issues • Informed consent • Confidentiality • Anonymity It must not be possible to link any response to a particular individual.

Anonymity • Use pseudonyms or numbers instead of names of informants or those named

Anonymity • Use pseudonyms or numbers instead of names of informants or those named by them • Do not attribute quotes in final report such a way that informants can be identified

Ethical issues • • Informed consent Confidentiality Anonymity The effect of the study on

Ethical issues • • Informed consent Confidentiality Anonymity The effect of the study on participants Will potential harm to the subjects outweigh the benefits of the study? • Quality assurance • Health and safety of researchers

Quality assurance • Particularly important for qualitative research studies • Personal views, sympathies and

Quality assurance • Particularly important for qualitative research studies • Personal views, sympathies and prejudices of researchers must be minimised • Validity: confidence in the accuracy of the findings and that they are reflective of the situation under investigation

9 Reporting and disseminating results A final report should: widen the knowledge base be

9 Reporting and disseminating results A final report should: widen the knowledge base be relevant to policy-makers and those who plan interventions

Structure of final report Executive summary Introduction Methods Results Discussion and recommendations

Structure of final report Executive summary Introduction Methods Results Discussion and recommendations

Executive summary 4 or 5 pages summarising all the information in the report in

Executive summary 4 or 5 pages summarising all the information in the report in short paragraphs

Introduction At least 3 pages to demonstrate the rationale for the study: • General

Introduction At least 3 pages to demonstrate the rationale for the study: • General description of location of research • Drug use context • Specific concerns of the study

Methods Shows how the aims and objectives were achieved: • • documents reviewed criteria

Methods Shows how the aims and objectives were achieved: • • documents reviewed criteria for inclusion in the study sampling procedures location of interviews / focus groups problems and successes research instruments data collection procedures data analysis process

Results Not the place for discussion: should present results only

Results Not the place for discussion: should present results only

Results • Characteristics of informants • Headings according to aims and objectives • Be

Results • Characteristics of informants • Headings according to aims and objectives • Be clear where results originated • Compare results from different information sources

Results How do you report sensitive issues?

Results How do you report sensitive issues?

Use of quotes Use to illustrate points made in the results section Too many:

Use of quotes Use to illustrate points made in the results section Too many: chaotic ‘stream of consciousness’ with no indication of why they are being used Too few: deprive the reader of rich qualitative data and the sense of contact with informants via their verbatim statements

Results Report ALL informants’ points of view

Results Report ALL informants’ points of view

Generalisability Do not over-generalise: results from a study conducted with schoolchildren from ONE school

Generalisability Do not over-generalise: results from a study conducted with schoolchildren from ONE school cannot be said to apply to all schoolchildren in the area, the town, or the country.

Discussion and recommendations • Examine major results • Discuss and interpret them in terms

Discussion and recommendations • Examine major results • Discuss and interpret them in terms of implications for policy and interventions • Ensure recommendations accurately reflect findings • Make recommendations for further research • Make recommendations stand out: boxes or bullet points

Dissemination of results to: • appropriate policy-makers • those who plan interventions • managers

Dissemination of results to: • appropriate policy-makers • those who plan interventions • managers of relevant institutions (such as schools, social workers, police, community workers) • drug services • social research institutions • those professionals who assisted with the study • all informants? • conferences and meetings • libraries • internet