Research Basics (With Example)

Aims

• Broad purpose of study ‘Researcher want to investigate the relationship between potatoes and sleep’

Hypotheses

• Directional: ‘Eating more potatoes makes people sleep more’ 

o Specific direction predicted 

• Non-Directional: ‘Eating potatoes changes the amount people sleep’

o No specific direction predicted – not sure what might happen

• TEST THE HYPOTHESIS –either supported or rejected

• Need to be sure IV changes the DV

• Need to be controlled – may not eliminate effect

• DOES ensure not effecting one condition of the IV – good design

·         Situational variables

o   Connected with reseach situation e.g temp, time of day

o   Controlled by standardisation only thing that differs between 2 conditions is the IV

·         Participant Variables

o   Connected to research Ps e.g intelligence, age etc

o   Controlled through experimental design

o   Randomly assigning Ps – controls P variables reduces bias

·         Other variables: Demand effects – Ps behave in  a certain way (single blind), Investigator effects: researcher behaviours and appearance making Ps act differently (double-blind)

• Measure of whether replications would produce similar results

• Test-Re-test method

o Ps complete test twice, different times

o Correlation coefficient – highly positive = reliable

o Effect of practive may be a problem

• Equivalent forms

o Overcomes practice effect

o Comparable versions of the test

• Slit-half method

o Internal reliability

o Items on test all measuring same ting

o Ps take test once

o Correlated on both halves of test 

o High correlation – internal reliability

• Inter-observer reliability

o Observer need to generate same results

o Same beh

o Training improves bias

• Replicated and repeated experiments

• Consistent measurements

• Must be high consistency

Valid = measuring what we claim to be measuring

• Internal validity

• External validity

Methods of Assessment 

• Face validity

o The ‘eye ball’ test

• Concurrent Validity 

o Comparing result – new versus old tests

• Predictive Validity

o Predict performance in future tests