Describing Data

Statistics are used to

  1. Describe data
  2. Make inferences about population
  • Once statistics are in, it is helpful to construct frequency distribution
  • There are different ways to graph data
  • Pie chart good for nominal data with limited number of scores
  • Bar graphs are used for nominal or ordinal data, with height of bar representing frequency (categorical)
  • Frequency polygon for internal and ratio-scale data
  • Histogram uses bars to display internal and ratio-scale data since scale values are continuous
    • Can see where or how people responded and can see outliers
  • Not how we present findings in psychology – avoid frequency distributions in lab reports

Case Studies

  • In-depth analysis of particular individual
  • Good source of inspiration, but should never be used as evidence
  • In-depth, longitudinal examination of individual, more qualitative
  • Focuses on in-depth description of case rather than statistical analysis
  • Uses a variety of other methods: naturalistic observation, interviews, questionnaires, psychological testing, physiological measure
    e.g., Book: Phantoms of the Brain
  • Commonly used to study rare conditions and can provide great insight

Benefits

  • Allows formulation of hypothesis and theories
  • Can provide detail and greater understanding
  • Study of abnormality can provide insights into normal functioning

Limits

  • Should not be used to provide support for theories – no control groups
  • Suffers from internal validity – no causation
  • Since based on individual, cannot generalise
    ↓ internal validity
    ↓ external validity
  • Plenty of opportunity to colour research based on personal bias

Systematic Observation

  • Careful observation of one or more particular behaviours to quantify behaviours
  • Less global, more interest in specifics
  • Hypotheses are made in advance
  • More focused on observational method
    e.g., Video and coding of child’s play, interested in sequence and order of play behaviours
  • To observe and code specific behaviour, quantified using coding system

Coding System – A description of a set of behaviour of how they are coded, with clear operational definition (need to be simple and categorised)

Issues in Systematic Observation

  • Reactivity is also a problem here, but time with cameras usually solves problem
  • Inter-rater reliability – coders need to agree on coding method
    • More reliable with more coders
  • Better to look at segments of behaviour across time

Naturalistic Observation

Naturalistic Observation – Researcher makes observations in a natural settings with summary and interpretation

  • Qualitative approach → inspiration of ideas
  • Wide array of information
  • In-depth description
    • High external validity
    • Low internal validity
    • There are concerns with biases and reactivity
  • Interpretation of results
    e.g., Jane Goodall and Chimpanzees
  • Focuses on naturally occurring behaviours and recorded in-depth
  • Researcher must immerse self into environment of study
  • Detailed notes, organise observations around themes

Issues in Naturalistic Observation

  • Whether to be participant in setting being studied
  • Non-participant observer – does not join group but observes
    • Participants may become aware of observation
    • Not as immersed in situation
  • Participant Observer – participates
    • Might over sympathise with participants
  • Weather or not to use concealment
    • Reactivity – When measure (observation) changes participant behaviour
    • Concealed observation might be unethical
    • People become used to being observed and may become used to situation
  • Defining scope of observation – what aspects to focus research on
    • Limited in what we can write down
  • Difficult, inconvenient, time-consuming

Observing Behaviour

  • Starting point for science; making a few observations
  • With observations and accumulation of knowledge, theories can be made
  • Variables can be qualitative or quantitative

Qualitative Research

  • People in natural settings
  • In-depth on relatively fewer people
  • Conclusions based on interpretations drawn by researcher
    • Global and exploratory
  • Often no hypothesis tested, description of observations

Quantitative Research

  • Examining specific behaviours that can be easily quantified
  • Less in depth and larger samples
  • Conclusions are based on statistical analysis
  • Specific and focused

∴ Qualitative research begins research and provides ideas, then moves into quantitative domain