prob stats c2

• Interview method


  Personal interview with the respondent (calls for a meeting, visits at home)
• Indirect interview method


  Uses telephone conversation to interview the respondents
• Common methods of collecting data

  • Interview method
  • Questionnaire
  • Observation
  • Test
  • Experiment
  • Registration
  • Use of mechanical devices
• Direct interview method


  • Appropriate if the needed information is minimal and the number of respondents is few (less than 30)
  • Inappropriate if the number of respondents is very large and they are living very far apart, considered very costly and time consuming
• Indirect interview method


  • Quite expensive if there are so many respondents
  • Some individuals consider it impolite and improper
  • Biased because people with no telephone cannot be included
• Types of Questionnaire


  • Guided-Response type
  • Recall Type
  • Recognition Type
  • Dichotomous type
  • Multiple Choice type
  • Multiple Response type
  • Free Response type
  • Rating Scale type
• Guided-Response type questionnaire


  The respondent is guided in making his or her reply
• Recall type questionnaire


  Questions that require the respondent to recall information
• Recognition type questionnaire


  Questions that require the respondent to recognise the correct answer
• Dichotomous type questionnaire


  Questions with a Yes/No response
• Multiple Choice type questionnaire


  Questions with multiple answer options
• Multiple Response type questionnaire


  Questions where multiple answers can be selected
• Free Response type questionnaire


  The respondent is not guided in giving their reply
• Rating Scale type questionnaire


  Questions that use a numerical or verbal scale for the response
• Empirical Observation Method


  • Obtaining data through seeing, hearing, testing, touching, and smelling
  • Can be participant observation or non-participant observation
  • Can be controlled observation in a laboratory
• Test Method


  • Widely used in psychological and psychiatric research
  • Uses standard tests due to their validity, reliability, and usability
  • Examples: Aptitude tests, IQ tests, Achievement tests
• Registration Method


  Obtaining data from government agencies like the National Statistics Office, Land Transportation Office, Department of Education, etc.
• Mechanical Devices


  • Used in social, educational, chemical, biological, medical, astronomical, space and atmospheric research
  • Examples: camera, projector, video tape, tape recorder, x-ray machine, microscope, telescope, barometer, computer, radar
• Sampling Techniques


  1. Determine sample size using Slovin's formula

  2. Random Sampling: Lottery method, Use of Random Number Table
• Slovin's formula


  𝒏𝒌 = 𝑵 / (𝟏 + 𝑵𝒆^𝟐)

  Where 𝒏𝒌 = sample size, N = no. of cases, e = margin of error
• Random Sampling


  • All members of the population have equal chance of being included
  • Applicable when the population is not classified into different clusters, sections, levels, or classes
• Sample size (𝒏𝒌) and no. of cases (N) are always expressed or rounded off to whole numbers
• The margin of error (e) is always expressed in %
• Random sampling


  1. Names of respondents written on paper, rolled, and placed in jar

  2. Respondents included are those whose names are randomly picked from the jar
• Using table of random numbers


  1. Assign numbers to each member of target population

  2. Compare last digit of random number to assigned numbers to select sample
• Stratified random sampling


  Population divided into different strata or classes, each class must be represented in the study
• Stratified random sampling example

  • Researcher wants to determine average income of families in a barangay with 3000 families distributed in 5 purok
• Calculating sample size for stratified random sampling


  1. 𝒏𝒌 = 𝟑𝟎𝟎𝟎 / (𝟏 + 𝟑𝟎𝟎𝟎(𝟎. 𝟎𝟓)𝟐) = 𝟑𝟓𝟑 families

  2. Calculating required sample size from each Purok based on population percentage
• Cluster sampling


  When the geographical area or target population is too large, clusters (groups) are selected randomly instead of individuals
• Cluster sampling example


  • Doctor wants to study correlation between smoking and death rate, selects 3 out of 13 regions randomly to include all residents of those 3 clusters
• Purposive sampling


  Respondents chosen based on their knowledge of the information required by the researcher
• Purposive sampling example


  • Researcher wants to study history of Town A, selects senior citizens as most reliable sources
• Quota sampling


  Commonly used in opinion polls, researcher meets a predetermined number of sample units based on certain characteristics
• Quota sampling example


  • Salesman surveys females in public places to gather information on most common hair shampoo used
• Convenience sampling


  Researcher selects most accessible respondents, e.g. using telephone interviews or own students, but this can be biased