- Количество слайдов: 21
Qualitative and Quantitative Data COLLECTION, COLLATION, ANALYSIS (AND DESIGN)
What is Re-search? • • • Searching is natural process. My dog can spend hours searching for her bone. Is that Re-search? What does the ‘Re’ signify? A search is need bound, whereas Research allows us to use knowledge to conceptualise. Research is the use of our preconceptions, bias, ideas, uncertainty, and ambiguities, etc.
Data collection at PRIA At PRIA, data is collected for: • Training needs assessments, undertaking planning exercises • Monitoring and Evaluation – Baselines, • Impact stories Data is mostly collected at three levels: • Institutional mapping • Household level • Settlement level • Individual experiences / perceptions
Qualitative & Quantitative inquiry Qualitative Quantitative Subjective. Focus on narratives Objective and measurable. Focus on numeric values Why, how, what! Focus is on richness of descriptions Who, where, when! Focus is on hardness and reliability Few cases – largely in depth Captures the individual’s perspective Less in depth, but across a larger number of cases with a focus on comparisions Unstructured or semi structured Fixed response options No statistical tests Analytical testing Seeks to explain and understand, less generalizable Seeks to generalise and show patterns / deviances In Depth Interviews (IDIs), Focus Group Discussions(FGDs), Participatory Research Appraisals, Case study approaches, Participant Observation Structured surveys – telephonic, mobile based or on paper.
Quantitative Research Understanding Sampling • What do the terms ‘Sample’ and ‘Population’ mean? • Sample is the group of people representative of the said population • Who will you do the research on? How many? What factors and variables will you take into account when collecting data? • Sampling effects the way you are finally able to analyse Why the fascination with numbers? Bias prevention • Requires a representative sample Increased generalizability • Requires a large sample
Quantitative Research Understanding Sampling Q. All of you are amongst the many employees of PRIA. I am an independent researcher conducting a study where I attempt to find out the average salaries of employees. -What is the population and sample for the study? There are others who may not be attending! You represent programmatic staff only! For this study, the population must represent ALL of PRIA’s employees. Only then must you choose a sample.
Quantitative Research Understanding Sampling – Simple Random Sample Q. All of you are amongst the many employees of PRIA. I am an independent researcher conducting a study in various NGOs, where I attempt to find out the average salaries of employees. Step 1. Decide on the number you want as your sample. Step 2. Assign a numeric digit to each person. Step 3. Put the numbers in a HAT! Step 4. Pick your sample and GO! Step 3(2). Use Excel’s Random function (RAND)
Quantitative Research Understanding Sampling – Stratified Sample Q. All of you are amongst the many employees of PRIA. I am an independent researcher conducting a study in various NGOs, where I attempt to find out the average salaries of employees, and disaggregate this based on factor(s) Step 1. Use one or more variables to classify your population - Male / Female distribution - Age distribution - Location -What are the variables I could take into account? Step 2 - Assign numbers and pick from each Strata Step 3 - Position - Pick randomly, as discussed - Qualifications / Educations before (if required) - And many more
Quantitative Research Tools of measurement and other factors Tools: • Surveys and questionnaires are the primary mode. Factors to keep in mind while designing: • Validity: Is the instrument being used actually measuring what it is supposed to be measure? • Can we use marks as a measure of intelligence? • Remember, correlation is NOT causation • BUT observation based methods can also be utilised. • How many women members participated in the Panchayat meeting? • Questions should be phrased carefully. They should not be leading, and they should take into account ‘social desirability’. Look at the following as bad examples: • Are you attracted to people outside your relationship? • Do you beat your children? Qualitative
Quantitative Research Understanding basic analysis पर व र म ल क जनम परम ण पतर क प स उपलबध प तरत दसत वज़ आध र क रड बक ख त घर क व टर क रड ककर ट (N=41) क छत व ल मक न 135 कचच परक र नह ह 35 85% 22 16 42 71 85 86 115 6 15% ह पकक मक न
Quantitative Research Understanding bivariate / crosstabulated analysis
Qualitative Research Why qualitative? And What’s different! Subjective quality of data – LOT to offer via stories, experiences and narratives Reality is constructed socially. As a result, probing and rambling need critical attention Go Qualitative when you do not know enough to quantify. Remember that quantitative tools have finite options! Qualitative research needs one to identify our values and bias because our questions and thoughts often flow from our own value systems and interests. Qualitative interviews conducted as ‘insiders’, where you maintain a closer social relationship with the participant.
Qualitative Research Understanding sampling Q. I am part of the Admin team at PRIA and want to find out what you think of the beverages available at PRIA’s offices. What are the various sampling methods I can use? 1. Convenience Sampling: ‘Convenient’ because those being studied are available at the time of sampling. However, I will miss out those that are not available in the room. 2. Snowballing: I will start by selecting one of the directors because they all drink a LOT of coffee. The directors will then connect me with others. Useful when a certain group is being studied, but has a risk element. In this case, who would the directors refer me to? 3. Quota Sampling: I will start by identifying the categories that are important to the study, and then picking a sample within them. What are the categories of beverages available at PRIA? 4. Purposive Sampling: Used to serve a very specific need or purpose. If I only want to know what people think about the coffee, and I know that is exactly what the directors consume, then it is them I will go to.
Qualitative Research Methods of data collection Participant and non participant observation Open ended or Semi Structured Interviews / Questionnaires Focussed Group Discussions Case study approaches
Interpretation, analysis and organisation Reflections Interpretation and analysis are not the end of the research process! Are the meanings you ascribe to people’s words and actions. “What does this mean”, “what am I hearing”, “how does this answer our question” People’s ramble, their expressions, hesitations, feelings etc can often tell a story Analysis & Synthesis To analyse means to take apart. To synthesise means to put something together. When you analyse information and ideas, you take it to pieces and then see what kinds of categories, patterns, trends, relationships can be developed
Qualitative Research Organising information - Two broad ways to organise data: - Qualitative data can either be organised through quantification, and the use of statements and quotes. Example: Five out of the six respondents interviewed liked the coffee, with one of them saying that he enjoyed the fact that he could control the amount of milk per cup, and that this was a facility unavailable in most machines. One of the respondents said it was too hot to drink coffee, and she was upset that cold coffee was not made available. - Or through the assigning of variables, themes and labels like gender, type of drink consumed, number of cups consumed, etc. - This process is called coding
Qualitative Research Analysing a case study Q. What constitutes a ‘Case’ - A ‘Case’ can be a person / people, groups, critical events, communities, settlements, etc. - A ‘Case’ can be either one piece of analysis, or an analysis could be made up of various ‘Cases’. Think about a settlement and how it fits into both these categories. - Data can consist of interview data, observational data, records, impressions and statements, diary entries, and any other information - A case is a bounded system. Either ‘time’ or ‘space – geography’ restrict it.
Qualitative Research Constructing Case Studies Analysing a case study Assemble the raw case data Construct a case record Write a case study narrative
Qualitative Research Understanding coding Step 1: Read. - Browse through all your data, irrespective of which method you utilised Step 2: Labels / Coding - In each separate transcript / case study / interview, start using labels for important words, phrases, sentences, sections and paragraphs - Labels can be for whatever you feel is relevant - Actions, activities, concepts, differences, opinions, processes Step 2. 5: Relevant? ! What’s relevant? - Something that’s repeated several times or assigned importance to by the interviewee Something that surprises you or is unexpected Something that connects with an article or report that you may have read Or something that reminds you of a theory or concept or contradicts one Something that shows underlying patterns Or anything else!
Qualitative Research Understanding coding Step 3: Re-label into themes - Bring the important and similar codes together into common ones Step 4: Look for other variables - Education Age Income And many more Step 5: Frequency count Step 6: Findings
Qualitative Research Pitfalls - Thematic analysis is not quotations! Use quotes as statements to substantiate - Data collection questions are not your themes - Do not get carried away. The analysis must remain grounded in the data