Cours5

• Cyberpsychology


  The study of how humans and computers interact
• Cyberpsychology research areas


  • Online behavior
  • Work behavior
  • School behavior
  • Etc.
• Measurement


  Specified procedure to convert an observation to a number
• Measure


  Reliability and validity
• Thermometer


  • Checking for a fever = take your temperature a few times to make sure the reading is correct
  • The values are really similar = the thermometer is reliable
• Reliability


  The consistency of the value obtained as you repeat the measurement
• Validity


  If the measures actually measure what is says it does
• Types of reliability


  • Test-retest reliability
  • Internal consistency (Cronbach's Alpha and McDonald's Omega)
  • Parallel forms reliability
  • Inter-rater reliability
• Cyberpsychology goal is to get as reliable measures as possible
• Example of invalid measure

  • You create a survey that asks a series of questions to check if the users like the app, but you phrase the questions awkwardly and the survey lets you really know about the person's current mood regardless of the app
• Scales used in cyberpsychology

  • Burnout
  • Turnover intention
• Eye movement tracking


  A common measure used in cyberpsychology to track how users view different interfaces and are drawn to different parts of a computer screen
• Other measures used in cyberpsychology

  • Mouse clicks
  • Mouse movements
  • Screen touches
• Measuring the impact of interacting with computers on broader psychological issues

  Often uses measurements scales like self-esteem or some form of satisfaction (e.g., life satisfaction)
• Observational study


  • How to capture actions
  • The Observer XT - record, code, analyse eye tracking, facial expressions, and physiology
• The Hawthorne effect


  If people are being observed, their behavior often changes
• Observing covertly


  Not always ethically possible
• Observer bias


  Specificity in recording observations might leave us not being able to record events of interest that might happen during the study
• Correlational designs


  The goal is prediction, we measure at least two variables and try to determine if changes in one variable are related to another variable
• The study was not an observational study because the researchers did not directly observe these behaviors
• External validity


  Generalizability of the study, can we generalize our findings?
• Factors influencing external validity


  • Population - not necessarily everyone (not everyone plays video games)
  • Population -> sample (a subset of participants)
  • How we identify the population is extremely important in the field of cyberpsychology - use and reach of computers is highly variable and depends on many factors
• The sample must represent the population
• Scale construction


  People are asked a series of questions to which they provide answers, using a Likert scale
• Likert scale example


  • Seeing photos on Facebook of my friends make me feel less happy

  Agree 1 2 3 4 5 Disagree
• Sentiment analysis


  The use of natural language processing (NLP), text analysis, computational linguistics, and biometrics to systematically identify, extract, quantify, and study affective states and subjective information