04 - Sampling and Data

Sampling and Data

Sampling and Data

• DataSet of values of qualitative or quantitative variables
• Lowest level of abstraction, from which information and knowledge are derived

• population: không gian mẫu, toàn bộ x (total clients, total guests)

• data: số liệu thực tế của các variable (1 giờ, 3 ngày, etc)

  • set of values of qualitative or quantitative variables
  • lowest level of abstraction
  • info and knowledge are derived from this
  • from sample data → calculate statistic

• variable: các measuring units của data (hours, minutes, pieces, …)

• statistics:

  • a number represents property of the sample
  • an estimate of population parameter
  • what you can obtained from the sample that can represent the parameter / goal of measuring

• parameter

  • numerical characteristic of the whole population, estimated by a statistic
  • goal of measuring / surveying in this case

• Attribute and Variable - “measurable?”

  • Attribute: characteristic of an object that cannot be measured
    • Example: sensibility
  • Variable: something that may or does vary and can be measured
    • Example: height

• Discrete and Continuous Variables - “precisely countable?”

  • Discrete: variable that can only take a countable number of values
    • Example: number of employees
  • Continuous: variable that may take on any value
    • Example: height of people

Types of Data Collection

• Census: data collection about everyone or everything in a population
  • Advantages: high accuracy
  • Disadvantages: expensive, time-consuming, out of date
• Sample survey: data collection from a part of the population
  • Advantages: less expensive, faster
  • Disadvantages: less accurate, depends on sample size and methods
• Primary and Secondary data
  • Primary: data collected directly for the purpose of the survey
  • Secondary: data collected for some other purpose, but can be used for the survey
• Methods of Obtaining Sample Data
  • Observation: gathering data by watching behavior, events, or characteristics in their natural setting
    • Advantages: understand ongoing process or situation, gather data on individual behaviors or interactions, know about physical setting
    • Disadvantages: data collection from individuals may not be realistic
  • Experimentation: testing competing models or hypotheses, or testing existing theories or new hypotheses
  • Qualitative techniques: investigating the why and how of decision making, using smaller but focused samples
  • Questionnaires: series of questions for gathering information from respondents
    • Advantages: quick, cheap, standardized answers
    • Disadvantages: may frustrate respondents, may lead to biased results
    • Methods: phone or personal interviews, postal surveys, self-completion

Sampling

• Process of selecting a sample of items from a population
• Why using sample instead of census: completeness, cost, time, accuracy
• Sampling frame: list of all those within a population who can be sampled
  • Characteristics: completeness, accuracy, up to date, non-duplication
• Random sampling: every item in the population has an equal chance of being included
  • Drawbacks: unrepresentative sample, scattered population (more spread out rather than focused)
• Quasi-random sampling (method of selecting samples combining random and non-random sampling): approximation to random sampling, includes systematic, stratified, and multistage sampling
  • Systematic: select an element from the list at random and then every kth element
  • Stratified: divide population into homogeneous subgroups (subgroup sharing similar traits) and then sample within each subgroup
    • Advantages: representative, precise
    • Disadvantages: not useful when the population cannot be partitioned
  • Multistage: divide the population into groups and then sample within selected groups
    • Advantages: cost, speed, convenience
    • Disadvantages: bias, not truly random
• Non-random sampling: used when the sampling frame cannot be established, includes quota and cluster sampling
  • Quota: segment the population into subgroups and then select subjects from specific subgroup, not of all population
    • Advantages: time, budget, accuracy
    • Disadvantages: unreliable, biased
  • Cluster: divide the population into groups and then sample all the elements in one or more groups
    • Advantages: convenient, practical
    • Disadvantages: less accurate, less representative

Survey Methods

• Two main categories: questionnaire and interview
• Two types of questionnaires: postal and group administered
  • Postal: questionnaire sent by mail to respondents
    • Advantages: inexpensive, same questionnaire, respondents’ convenience, time to read
    • Disadvantages: low response rate
  • Group administered: questionnaire filled by respondents in a group setting under supervision
    • Advantages: higher response rate
• Interviews: may be qualitative or quantitative
  • Quantitative: personal and telephone interview
    • Personal: face-to-face conversation between interviewer and respondent
      • Advantages: high response rate, low response errors
      • Disadvantages: time-consuming, expensive, simple questions
    • Telephone: voice conversation between interviewer and respondent
      • Advantages: rapid, no travel, sensitive questions
      • Disadvantages: high refusal rate, short interview
  • Qualitative: focus group
    • Focus group: group of people asked about their perceptions, opinions, beliefs, and attitudes towards a product, service, concept, etc
      • Advantages: data and insights from group interaction, common language
      • Disadvantages: scheduling difficulty

Questionnaire Design

• Basic rules for questionnaire construction:
  • Each question should be clear, unambiguous, and easy to understand
  • Every respondent should be able to answer every question
  • Each question should relate directly to survey objectives
  • Question should not be biased or make assumptions
  • Do not use double-barreled, long, negative, or guessing questions