3067 MC Research Methods John Keenan john keenancoventry

306/7 MC Research Methods John Keenan john. keenan@coventry. ac. uk

Primary Research – original done by the researcher Secondary Research – found by the researcher from other sources

YOU DO NOT HAVE TO DO PRIMARY!

If you do. . . read first http: //www. socialresearchmethods. net/kb/index. php Books: 302. 2 Media research methods Blog link for Power. Point: http: //whisperingchicken. wordpress. com/

306 MC If you choose it your likely dissertation structure will be: 1. Introduction – outlining dissertation (what, why, how) 2. Literature Review – previous studies, gap in knowledge 3. Methodology – how you will study topic 4. Results – what you found 5. Conclusions – connect literature review with findings

307 MC No structure Present results in portfolio

Validity - does what it should Reliability - degree to which measurement is free from error

Validity Ask the right people Ask the right amount of people Ask the right questions “I like this product the best because it didn’t make my dog as sick as the previous four” http: //www. googleduel. com/reportindex-popculture. php

Reliability the same results gained over re-testing Re-write the questions to test reliability eg would you drink beer out of a bucket? Is a bucket a suitable container for beer? Re-test in same way

Qualitative Research Advantages: §does not impose reasons §gets understanding not confirmation §gains depth of knowledge §can see the world in its natural state §philosophy of the free-thinking individual §Discovers insights

Qualitative Research focus group – 8 -12; recorder; moderator Interviews – depth; interview schedule; probes; record participant observation/field – record case study Systematic replicable

Why participant observation? Preliminary study Inexpensive Natural setting – Hawthorne Why not? Validity – what is recorded, one viewpoint Ethics – Covert? (if not, Hawthorne) Danger – Street Corner Society Data Cameras, sound recorders – transcribed Notepad Purpose led Give subjects cameras; complete diaries

Why focus groups Speed Interaction Why not? Monopolisation Inhibited Recording Record-transcribe Non-verbal?

Why interviews? Depth Easier Why not? Interviewer bias Recording Record-transcribe

Why case study? simple natural Why not? time consuming not generalisable Recording Record-transcribe Notes, diary

Data Discourse analysis Programs you can use: http: //www. pressure. to/qda/

Sampling for qualitative research Purposive Representative

Quantitative Research Measured Advantages Provides evidence - facts and figures Can be re-tested Can be generalised Disadvantages Gives figures not reasons There are lies, damn lies and statistics 82. 3% of statistics are made up on the spot

Types of quantitative research Questionnaire Survey Content analysis Experiment

population entire group of people or media about whom information is needed

Types of Sampling Convenience Quota sampling - proportion Random - eg 1 in 10 of population Snowball - contacts find contacts

Structure of Questions 1. Screen - find target audience 2. Warm-ups - easy-to-answer questions 3. Increasingly difficult questions 4. Classification (if not in 1)/personal questions

Screen Ask key questions first e. g. to get home-makers Bought a house? Got children? Eg to get earth-mothers How important is recycling to you What are the most important things about a car?

Problems to avoid 1. personal How many sexual partners has your mother had? a) 5 b) 6 c) 7 d) more 2. Leading 3. difficulty “So, Debbie, what first attracted you to the millionaire Paul Daniels? ” When did you stop beating your wife? 3. Easy options Pilot it!

Classification questions gender, age, marital status, class, occupation whatever is necessary Please tick Age: Under 20 21 -24 25 -39 40+ Gender ethnicity M W F B A C O

Questionnaire/Survey Design Measurement Scales nominal - numbers How often do you visit the pub in a week One Two Three Four Five Six Seven likert scale - statements possibly from qualitative each item is classified beforehand eg I always send my wife to get the drinks All I want from the pub is to sit down and drink as much beer as I can

Questionnaire/Survey Design Measurement Scales graphic - pictures or line eg how comfortable are you with the idea of censorship?

Questionnaire/Survey Design rank order eg what is most important about beer - please rank - 1(most important) - 5 (least important) taste___ price___ brand___amount___ accessibility____ semantic differential scale eg how does drinking beer out of a bucket make you feel proud________________________ashamed wild_________________________ controlled happy ________________________ sad

Methods of Data Collection street in-home place of work venue phone post internet

methods used to analyse images • semiotic analysis • content analysis Images are not neutral Text anchors images

• population • sample • construct categories of content • establish a quantification system (coding schedule) • pilot • code • analyse data

Anchorage This is the last painting by Van Gogh before he killed himself

reliability

reliability

reliability

validity mother • synonyms for mother • referenced to female parent • referenced to term of abuse • in a child’s voice • spoken by parent

Experiments Correlation between variables Independent variable (predicts a change) dependent variable (will change) independent affects the value of dependent

Textual Analysis Methods Discourse Analysis http: //bank. rug. ac. be/da/da. htm Structural analysis - Claude Levi-Strauss http: //www. colorado. edu/English/ENGL 2012 Klages/levi-strauss. html Narrative analysis - Vladimir Propp http: //social. chass. ncsu. edu/wyrick/debclass/Propp. htm Psychoanalysis - Lacan www. colorado. edu/English/ENGL 2012 Klages/lacan. html Semiotic Analysis Content Analysis

Dealing with Data • Tabulation • Average: mean, mode, median • Standard deviation • Correlation

Statistics transferred to program Excel SPSS

Percent frequency count % Significance?

Average mean - add all divide by total median - line up all responses and find the one in the middle mode - find most popular response Standard deviation square root of the variance how much the scores differ from the mean

Correlation Analysis relationship between one and the other EXCEL - CORREL

ethical issues