Friday, 24 November 2017

planning + conducting research


  • Aims + hypotheses
Research aim = points out what the research is meant to accomplish and states the purpose of the investigation

Research question = asks about what the study aims to investigate

Null hypothesis = predicts no difference or no relationship between the variables

Alternative/2 tailed hypothesis = predicts a difference or relationship between the variables but doesn't predict what it will be 

1 tailed hypothesis = predicts the outcome or relationship between the variables

  • Sampling methods
Target population sample = the group the searcher is interested in

Random sample = each member had an equal chance of being chosen (very representative)

Snowball sample = people who are easily available chosen, then they bring friends (not representative)

Opportunity sample = people who are around and easy to choose (quick and easy)

Self-selecting sample = volunteers (not representative)

  • Experimental designs
-Repeated measures - each p takes part in every condition

+ fewer p's needed and individual differences are controlled 

- p's may suffer from order or practice effects leading to poorer performance

Practice effects = getting better at a task with practice and performing better

Order effects = getting bored of a task and beginning to perform worse

These can both be avoided by splitting each group in half and making them do the conditions in different orders 


-Independent measures = different p's used for each condition

+ no chance of order or practice effects

- more p's needed

-Matched participants = p's who are similar being placed in each condition, e.g. matching them on aggression levels

-may be hard to find matches


Single blind = p's unaware of the IV, so less demand characteristics 

Double blind = researcher also unaware, so no researcher bias

Mundane realism = extent to which findings can be generalised to realistic situations in everyday life


  • Designing observations + self reports

Behavioural categories = categories used to break streams of activity into recordable events

Coding frames = the categories put into codes, e.g. walking/talking/laughing

Time sampling = recording pre-determined behaviours at intervals, e.g. every 15 seconds

Event sampling = counting number of behaviours in time periods, e.g. tally charts

Open questions = allow detailed answers in p's own words and produce qualitative data, but are hard when identifying patterns or making comparisons

Closed questions = explicitly stated responses to choose from which produce quantitative data


  • Rating scales
Numerical scales that produce quantitative data

Likert rating scale = allows p's to say how much or to what extent they agree or disagree

Semantic differential scale = gives choices, and you choose 1 only

+ easy to analyse

- subjective and not always valid

Share:
© Psychology OCR notes | All rights reserved.
Blog Layout Created by pipdig