PSY247 Chapter 1,2,3,5,6,7,8,9,10,11,12,16,17: PSY247 Research Design and Analysis Textbook Notes
PSY247 READINGS
CHAPTER ONE –
Statistics – is the study of how to collect, organise, analyse and interpret
numerical information from data
Inductive statement – statement whose truth is assessed by collecting and
analysing data
• The term truth has different meanings depending on its context
• Inductive statements are based on statistical reasoning
• Example: the more hours you spend in the library, the better your
academic performance
Statistical Reasoning: Tools used to evaluate inductive statements
Behavioural science has developed 2 specialities that are part of the rational
inductive process:
1. Research design
2. Statistical reasoning
Research design – the science of collecting data, making observations about the
real world, considering how many observations to make and under what
conditions to make them
Statistical reasoning begins with the collected data and prescribes the rules by
which rational statements about those data can be made
Research design and statistical reasoning are intertwined
Pygmalion effect (also known as experimenter bias): people act in accordance
with others’ expectations
• People tend to do better when treated as if they are capable of
success
• ‘You can get what you expect’
• Observed by Eliza Doolittle & George Bernard Shaw
Psychologist Robert Rosenthal studied if the Pygmalion effect exists in different
settings
• Pygmalion effect applies in the rat lab – rat handlers expect quick learning
→ they get quick learning
Oak School – administered ‘Harvard Test of Inflected Acquisition’ (HTIA) which
was actually a standardised test of intelligence called Tests of General Ability
(TOGA)
• Intellectual growth – post-test IQ – pretest IQ (+ve scores suggest
increase in IQ)
• Pygmalion effect exists in the classroom
The Pygmalion Effect, known as experimenter bias, is now an accepted fact that
can affect research resign. To combat, research studies can use:
Placebo – in a drug study, a substance that looks like the drug being tested but
actually has no effect
Blind Design – not knowing which subjects are assigned to which experimental
condition
Single blind → either the participants or the research does not know
Double blind → both the participants and the research does not know
The experimental/statistical examination of the Pygmalion effect → led to
important social consequences
This example (drug study) shows 3 characteristics of the
experimental/statistical method:
1. The inspiration for an experiment is a pre-experimental observation
Example: Shaw observed characteristic of human nature and Rosenthal and
Jacobson seek to verify observation
2. Statistical tools must distinguish between samples and population
3. It must allow us to derive meaningful results from data that are not
perfectly consistent –that fluctuate from one person/situation to another
Statistical toolbox must include methods to measure the differences b/w
individuals
Samples from populations:
Population – all the members of the group under consideration
Sample – a subgroup of a population under consideration
Statistic – refer to any measurement on a sample – e.g. sample mean → statistic
Parameter – any measured characteristic of a population – e.g. population mean
→ parameter
Sample statistics are typically used to estimate population parameters
• Descriptive statistics → to describe our samples
• Inferential statistics → generalise from our samples to the wider
population
Measuring all the elements of a population is:
• Costly
• Difficult
• Impossible
Hence, samples are drawn b/c:
• Cheaper
• Quicker
• More convenient
1 of the most important questions in stats: What can we say about a population
when all we know about is a sample?
Types of Sampling:
Random sample – every member of a population has an equal chance of being
elected
• Unbiased representative sample BUT time-consuming
Stratified sampling – dividing the target population into important
subcategories, and selecting the members who have the most characteristics that
represent the population
• Time-consuming
• Large effort to make sample representative
Volunteer sampling (also called self-selecting) – individuals who have chosen to
be involved in a study.
• Participants receive an incentive
• Can be unrepresentative
Opportunity sampling – simply selecting those people that are available at the
time
• Quick and economical