RM

• Research methodology


  1. Problem statement

  2. Research question and factors

  3. Sub-research questions

  4. Conceptual model

  5. Hypotheses

  6. Operational and conceptual definitions

  7. Indicators
• Unit of analysis


  The level at which the research is performed, and which object are researched
• Population


  The group we want to say something about
• Sampling frame


  The group you can say something about
• Sample


  Your selection from the sampling frame
• Characteristics of a correct sample

  • Accuracy: no systematic bias
  • Precision: values in the sample like values in the population
• Sampling error


  The values in the sample may not be sufficient like the values in the population
• Sample types


  • Probability sampling
  • Non-probability sampling
• Probability sampling


  • Based on random selection, each unit has the same chance of being selected
• Probability sampling types


  • Simple random sampling
  • Systematic sampling
  • Stratified sampling
  • Cluster sampling
• Population parameters


  Summary descriptors (mean, variance) of variables of interest in the population
• Sample statistics


  Descriptors of the relevant variables computed from sample data, used as estimators of population parameters
• When the variables of interest are measured on nominal or ordinal scales, we use the sample proportion of incident to estimate the population proportion and the pq to estimate the population variance
• Population proportion of incidents

  The number of elements in the population belonging to the category of interest, divided by the total number of elements in the population
• Systematic sampling


  • Every Kth element in the population is sampled, the Kth element is determined by dividing the sample size into the population size
• Stratified sampling


  • The population is segregated into several mutually exclusive sub-populations or strata, the sample is constrained to include elements from each of the segments
• Ratio data


  Order in numbers is important, intervals between numbers are fixed, meaningful zero point
• Interval data


  Order in the numbers is important, interval between numbers is fixed
• Ordinal data


  Order in the numbers is important, intervals between numbers are not fixed
• Nominal data


  Numbers indicate the categories, order is not meaningful
• Avoid nominal variables, only use for controlled variable such as education
• Response method


  Determines the data type, not the concept itself
• Stick to 5-7 answers, answers need to be mutually exclusive
• Double-barreled question, never ask two questions in one
• Symmetric scale


  Treat as interval
• Asymmetric scale


  Treat as ordinal
• Good survey question


  • Clear and precise, can be answered meaningfully by the respondent, provides informative data
• Avoid asking sensitive questions if you do not need to, make sure the answers are mutually exclusive, categories should be exhaustive, no vague questions, use the highest data that makes sense, do not involve the option 'don't know' in a factual question, make sure timing period is involved
• Creating correct survey questions

  1. Should this question be asked?

  2. Is this question clear?

  3. Can the participant adequately answer the question?

  4. Does the participant want to answer this question?
• Examples of what is wrong with survey questions

  • Leading question
  • Ambiguous question
• Avoid ambiguities: use informal language, be precise, avoid double negatives
• Positive test strategy


  Participants have the tendency to answer in a positive way
• Question order


  • Funnel approach: from general to specific, group questions and introduce each set
• Open questions


  Difficult to work with, some people give long stories, some give few words
• Closed questions


  Limit participants in what and how they can answer
• Closed question types


  • Dichotomous scale
  • Categorical scale
  • Likert scale
  • Semantic differential scale
  • Numerical scale
  • Graphic rating scale
• Ensure that the answers you use in the scale are contradictions, easiest to state as 'the more the better', balance, be specific
• Ranking scales


  • Paired comparisons
  • Ranking scales
  • Constant sum
• Simple random sampling


  Each unit has the same probability of being selected (marbles in a vase)
• Simple random sampling


  • Often impractical, requires a lot of information about your sampling frame, may result in samples that do not fit your research question