RMA - Week 2

Clever Hans study

• Was believed that a horse could count

• Turns out he was responding to physichal cues (non-verbal)

• Researcher unconsciously influenced participants

Cofounding variables

• an outside influence in research that changes the relationship between the independent and dependent variables, creating false correlations or skewed results

Non-experimental methodologies

• Descriptive research

• Naturalistic observation

• Case study

• Survey

• Correlational research

Descriptive research

• naturalistic observation

• Case studies

• Surveys

naturalsitic observation

• categorisation

• little disturbance to natural processes - e.g. covert observation

• natural environment - with slight manipulation

Issues and solutions : observations

I - Lack of reliability with categories

S - 2+ researchers compare notes - inter-observer reliability

I - reactivity - change of natural resposne due to observation

S - participant observations - training a passing person in environemnt to observe e.g. parent of child

case studies

• observation of one partiular person

• usually a unique/ rare case

• valuable but should be approached with caution

Issues & Solutions : Case studies

I - Generalisation - results may only be applicable to one or few people

- Reproducibility - can results be recreated

-Cause-effects - How can we determine that the variable is actually causing a specific outcome

S - Similarities of cases with some differences - testing what staays constant and what changes

Surveys

• Questionnaire

• Interview

• Diary study

Issues : surveys

• Reactvity - participants aware that results will be analysed

• Questionnaire validity - does it test what it intends to

• Comparing answers - words on a scale could be interpreted differently

• Memory of participants - selfreport may be flawed by false memories - bias must be acknowledged

Structured interviews

• Fixed set of questions asked in a fixed order - no elaboration

• Same questions for all participants

• multiple choice/ ratings

S/W : Structured interviews

S

• Easily quantified/ analysed

• Comparable

• All topics covered

W

• Rigid structure

• Not adaptable to each pps - Not aware of greater context of answers given - only surface level

• Surface information

Unstructured interviews

• number of topics with no fixed order

• Flexible

• Can change based on conversation at hand

S/W - Unstructured interviews

s

• In-depth info

• Relevant to specific participants

w

• less generalisable

• time consuming analysis - e.g. thematic analysis

S/W - Descriptive research

s

• someitmes only way to study something

• inexpensive / flexible

• real life - no manipulation

• eclogically valid

Correlational research

• used to determine relation between two variables

• no manipulation of variables

• Quanitifable variables

Random distribution

no relationship

S/W - Correlational research

S

• degree of relation stated - identify quality of relationship

• prediction is possible

W

• direction of relationship unclear

• cofounding variables