Ryan Ward Block (Yapper)

Tools

Summarising and describing data you’ve collected

Theory

Math behind rules and tools, how experiment reflects real world

Dualism

The idea that mind and body are separate

Four Goals of Science

Description, Explanation, Prediction, Control

Authority Approach

Getting information from sources thought to be valid and trustworthy

Analogy Approach

Applying existing similar knowledge to new situations

Rule Approach

Try to establish laws or rules that cover a variety of abservations

Empirical approach

Testing ideas against actual events - Observing behaviour and drawing conclusions

Experimental Research

Experiments involved with explaining what caused behaviour or controlling a behaviour

Correlational research

Experiments involved with describing and or predicting behaviour

Descriptive statistics

Summarising the data collected from a sample

Inferential Statistics

generalise results from the sample to the population

Operational definition

Measuring a dependant variable through a value that indirectly reflects the property of interest

Nominal Scale

Categorises without ordering, numbers that substitute for names, for unordered categorical data

Ordinal Scale

Categorises and orders the categories, bigger means more though can’t tell how much more. Distance between points not considered equal

Eg: Rugby team standings, rank order couples in amount they love each other

Interval Scales

Categorises, orders and establishes an equal unit of measurement in the scale, knowing how much more, distance between points considered equal eg; celsius temp

Ratio Scales

Categorises, orders, establishes an equal unit in the scale and contains a true 0 point, allows ratio statements: “twice as big” eg: # of items recalled in memory task

Pepsi Challenge

Shows unexpected confounding variable, 85% chose S cup over L cup

Manipulated IVs

Experimenter manipulates variable, prediction and explanation, is a True experiment

Subject IVs

Recruit people in different groups, prediction but not explanation, is a Quasi experiment

Quasi Experiment

May not assume there is a causal explanation between pattern of results and IV level

Control group

Looks at the dependant measure in the absence of any experimental conditions

Yerkes-Dodson Curve

The U shaped relationship between stress levels and performance

Demand characteristics

Cues in a new situation people interpret as ‘demands’ for a situation

Between Subjects

Each subject experiences one level of the IV, Subject variables may be confounds so need to randomly assign

Control variable

Could make a confounding subject variable this to increase control (Eg only test females), does reduce generisability

Within Subjects

Each participant experiences every level of the IVs, participant serves as own control

Practice Effect

Order in which participants experience IV can be problematic in within subjects

Counterbalancing

Each IV condition is equally exposed to the practice effects and demand characteristics inherant in the within subjects design

Fully Crossed

When collecting data with multiple IVs all combinations of IVs is collected

Factorial Designs

When a study with multiple IVs is fully crossed, Eg: a 2×2 factorial matrix or 3×3 factorial matrix

Mixed design

When one IV is within subjects and the other is between subjects

Main effects

The effects of one IV on the DV ignoring the other IVs in the study for instance averaging the ejnjoyment of the food wo acknowledging sauce

Inferential Statistics

Stats where you can infer changes in your data to the population

Descriptive stats

Stats that describe, variability itself in this context is interesting as is a property of the data

Variance

(S)²

Standard deviation

Is the approximate average of the scores in a data set from their mean

Inflection Point

Point on distribution where curve starts bending out, corresponds with 1 sd from the mean

Z-score

Tells us how far away score is from mean expressed as # of sd from mean (x_i -`x)/s

Pearson r

A numerical way to compute linear correlation

Curvilinear

Increase in x initially results in increase in y then decrease in y

Directionality Problem

Unsure as to whether x influences y or y influences x

Cross lagged panel correlation procedure

Using a follow up study measure correlation across time

alpha level

The probability value that defines the boundary rejecting or retaining H₀ also known as significance level

Null hypothesis

Predicts no relationship between variables

Alternate hypothesis

Predicts there is a relationship between variables

Type 1 error

Rejecting H₀ when H₀ is true

Type 2 error

Retaining H₀ when H₀ is false

Increasing alpha

Reduces probability of type 2 error, increases the power of (1-B, increasing the probability of rejecting H₀ when it’s false), increases probability of type 1 error