Title | Föreläsning 2 - Forskningsdesign, endast engelska anteckningar. |
---|---|
Course | Introduktion till vetenskapligt arbete |
Institution | Högskolan Dalarna |
Pages | 3 |
File Size | 56.7 KB |
File Type | |
Total Downloads | 95 |
Total Views | 118 |
Forskningsdesign, endast engelska anteckningar....
FÖRELÄSNING2, TISDAG 20191008
What is research design? Broad - the whole package of research, everything related to research Narrow - primarily the setup of the data collection process
A research design is… What to i want do understand? Causal connections between variables Generalization from sample to population Understanding phenomena in their specific context Determining how and why phenomena change Determines the method to collect the data
Research Typologies -
-
The Big Three - Experimental: Measuring the effect of change in one or more variable. Small group of people of the population of interest. One control group (constant variables) and one with manipulated variables. (Low lighting example) hypothesis testing. - Survey: S tandardized collection of information. Survey is a standardized form of questiones. NOT the same thing as a questionnaires. Can be interviews too. - Case Study: Detailed study/knowledge about a single case or a few. Using a lot of different methods, ex observations, interviews, documents. Triangulate the case. Other types - Cross-section - at one point in time - critic: what variable is actually affecting the result? - Longitudinal - over a period of time - advantage: we know how things change, cause of effects. - panel s tudy: often a random group of population, and data collected across a number of timepoints. - cohort study a group with a shared characteristic (date of birth, marriage etc.) across a number of timepoints - trend study: different individuals sampled from the same population (not the same people) and data collected across a number of timepoints not looking for change in individuals - Prospective - follow up (how change affects) perspective - critic: - Retrospective - look back (cancer example) what is the influencing factor - critic: no control - Comparative - two similar phenomena in different contexts - comparing two situations/cases contrasting - understanding the phenomena
Research strategies and methods Two major research strategies for how one should collect and analyse data - Quantitative - numerical information - theory testing: deductive - natural scientific model - social reality objective - large study sample: reliable representation - there is a common truth - Qualitative - text and language - phenomenological model: we construct our reality - small study sample: understand the world people see - a research project will usually follow one or another - alt. a mixed methods, more used now. use one to understand the subject even deeper - antagonistic
Research methods Techniques for collecting and analysing data - Sampling: T hree main aspects - Who: define a study population and a target population - How: random, purposive and convenience - How many: n umber of participants - Data collection - Interviews - focus groups - Questionnaires - open or closed questions; tests and scales - Observations - content/frequency of behavior; tests, scales - Documents - death certificates; diaries - Secondary Data - studies already made
Data Analysis Form of analysis tied to form of data - Statistical analysis (numerical) - description of variation - correlational analysis - do things vary together? - comparing groups - do they differ? - statistical significance - is this a chance finding? - Content analysis ( text or non-numerical) - description of context - what is said and done? surface process - what is meant? underlying process
Choosing a research design -
What is the purpose of the research?
Exploratory - find out what is happening, build theory. refine questions, not produce definitive answer - Descriptive - accurate profile of persons, events, situations, or pre-prepared questions to collect required information - Explanatory - seek answer to pre-specified questions. determine relationship - Population - quantitative methods - Sample - qualitative methods What is the research question? - should precede the design of the research study - the formulation of the question can imply aspects of the design required - decide on a research area - select a topic or issue within the research area - develop several alternative research questions - the questions should be articulated as a logical argument about observable/measurable things, events, people - Qualitative question: s hould not specify relationships or effects. what is it like? how do things occur? - Template for qualitative research question - How or what is the... - story of/account of - meaning of/perception of - theory that explains - issue of/experience of - Quantitative question: c an form the basis for an hypothesis (predictions about the nature) - does the theory or model explain the relationship between IV and DV What are the pragmatic restraints? - Expertise: skills - Resources: funds, equipment, time - Access: reach the case, sample? - Control: c onstrain certain factors to examine those of interest - Ethics: morally defined -
-
-
A good research question -
logical/simple expressed the question cannot allow for only one answer too broad a question or too many questions justifiable
Is it researchable?
capable of being answered? resources? will the answer be of value? answerable within a moral framework is it discriminatory?...