psyc 102 - 1

Week 1 - Methods

Empiricism

Explaining Other’s Behaviours

• Authority: an expert telling you information that you accept as true

  • External 

• Intuition: having a feeling something is true

  • Internal 

• Observation: knowing because you have experienced it 

  • Balance between internal/external (trusting your senses + internalizing)

• Observational problems:

  • Not always applicable, our senses aren’t perfect, and people can disagree

  • Observer bias: Expectation can influence observations/reality (fast vs slow rat study)

• Empiricism: used to describe the conviction that accurate knowledge of the world can be acquired by observing it

Solving Observational Problems

• Scientific method: procedures/rules for how we observe the world + fix observational problems

• Theories: explanations of why something works based on observations

• Hypothesis: predictions (if something is true, then __ should happen)

• Experiments: testing hypothesizes to make a theory

Ways to Fix Bias

• Scientific skepticism: not getting attached to a prediction

• Peer review: after you make a study, you send it to other scientists in the field to check it

• Replication: multiple studies should produce similar data

• Random sampling: selecting varying participants that have an equal chance of being included in the sample

• Openness: all data should be public

• Double-blind experiment: person collecting data/participant should not know the hypothesis

• Falsifiable hypothesis: an explanation that can be true/false with observation

  • If this then that

Methods of Observation

Scientific Method Steps

Hypothesize: 

• Confirmatory studies: start with a hypothesis and find data that proves it

• Exploratory studies: start without a hypothesis and find lots of data to come up with one

Operationalize: 

• Operational definition: a description of a psychological property in measurable terms

  • eg Can’t observe stage fright directly, need to measure something specific (heart rate)

  • Need construct validity: feature of operational def. that are good indicators of the property (eg smiling = happiness)

Measure:

• Instrument: tool that measures operational definition (questionnaires, computer tasks, physiological measures)

  • Qualities of a good instrument

    • Internal validity: measures what it claims to

    • Reliability: similar each time

    • Power: detects small differences

Definition + instrument = data

• Process of measurement:

  • Define the property (operational definition) → detect property (with a reliable instrument)

Report:

• Peer review: receiving feedback from other scientists

  • Skeptical, check instruments, check data

Assumptions:

• The scientific process needs necessary assumptions

Conclusions:

• Internal validity: attribute of an experiment that allows it to establish causal relationships (everything inside the experiment is working how it is meant to)

• External validity: attribute of experiment where variables can be described in a way that represents the real world

• Types of Error:

  • Type 1 👎➕ concluding there is a causal relationship when there is not - false positive 

  • Type 2 👎➖ concluding that there is not a causal relationship when there is - false negative 

Picturing Data

• Frequency distribution: graph showing number of times a measurement occurs on each value (bar/line graphs)

• Distributions: 

  • Negatively skewed = leaning right

  • Positively skewed = leaning left

  • Normal distribution = symmetrical and “bell curved”

• 3 descriptions of central tendency 

  • Mean: average value

  • Median: value in the middle of data

  • Mode: most frequent value

• Variability: wideness of distribution

  • Range: largest value minus the smallest value

  • Standard deviation: describes how values in the frequency distribution differ from the mean

Data Collection Methods (types of studies)

• Naturalistic Observation: gathering information by uninteruptingly observing their environment (some things cannot be observed this way)

  • Demand characteristics: people acting differently in a setting where they know they are being watched/expected to do something

  • Ways to avoid demand characteristics:

    • Privacy - People feel more comfortable when they are anonymous 

    • Unawareness - People being observed are unaware what they are being observed for

    • Measuring things people can’t control (pupils dialating)

• Case studies: Procedure for gathering informationn by observing one individual 

• Correlational studies: A numerical prediction of relation between two measured variables (Is there a relationship between ___ and ___?)

  • Measure the variables in many subject, graph the data → estimate “direction” and “strength” (correlation coefficient)

    • Direction:

      • Positive correlation: the value of one variable increases, and the other variable increases

      • Negative correlation: the value of one variable increases, and the other variable decreases

      • No correlation: no pos or neg relationship

  • Correlation ≠ causation

    • Directionality problem: for any correlation, we don't know if one variable caused the other or if the other caused the first (a might have caused b, or b might have caused a)

      • Solution: 

    • Third variable problem: a third unmeasured variable may be the cause for the measured one

      • Eg Ice cream sales and drowning incidents are correlated, but a third variable (temperature) influences both: hot weather increases ice cream consumption and swimming = more drowning incidents.

      • A situation where it seems like there is a causal relationship but isn't - Spurious correlations: strongly correlated variables that we know are not causally related

      • Solution: 

• Experiments: One variable is manipulated to see the causal effect on the other

  • Control group: doesn't receive the treatment being tested (not allowed to sleep)

  • Experimental group: does receive treatment being tested (can sleep normally)

  • Components of experiments:

1. Independent variable: the variable manipulated (eg amount of sleep)

2. Dependent variable: the variable measured (stress)

3. Random assignment: randomly assigning participants to groups

4. Random selection: should be representative of the population as a whole

• Meta-analysis: combining data from (5-20) other published studies to get a more precise result

  • Averaging studies

Ethics of Science

• Ethics code for the American Psychological Association:

  • Informed Consent: verbal agreement to participate in the study once knowing all risks involved

  • Freedom from coercion

  • Protection from harm: choose the safest methods

  • Risk-benefit analysis: participants may need to accept small risks but overall the benefits need to be grate

  • Deception: only when extremely necessary for the experiment

  • Debriefing

  • Confidentiality: anonymity

Respecting animals 

• CCAC Three Rs Tenet

  • Replacement: no alternative to animals

  • Reduction: using the smallest number of animals possible

  • Refinement: procedures need to minimize discomfort