research methods

• directional (one tailed = direction of results is predicted

  non-directional(two tailed)= a change or difference is predicted but not direction is specified
• repeated measure design= one one group of participants, take part in both conditions

  😃 more economical as less participants required

  😔order effects may take effect- feel fatigued or bored when it comes to the second condition
• independant group= two separate groups of participants for each condition

  😃less order effects- only have to do it once, no demand characteristics

  😔 participant variables, range of different participants
• matched pairs= two separate groups but they are matched into pairs for certain demographics

  😃 no participant variables or individual differnces

  😔 participants can never truly be matched even with twins, time consuming and drains resources
• order effets

  • occur in repeated measure design
  • practice effects- well practices to complete the second
  • become bored and fatigued after first condition
  • overcome by counterbalancing and randomisation

  -counterbalancing= half do condition b for a and vice versa, first and second condition not the same for all participants

  -randomisation= assigned to condition a and b by random generator
• random sampling= random generator, equal oppotunity

  😔 time consuming

  😃 free from observer bias
• volunteer sampling = advertised

  😃 easy to gather participants

  😔 volunteer bias, attract a certain type of person
• opportunity = list of participants all available then ransom allocate

  😔 may be unrepresentative as drawn from small sample

  😃 convenient saving time and money
• stratified= randomly select from sub-strata of participants

  😃 avoids research bias

  😔not reflect all participants
• systematic= list ion participants then select event nth number

  😃 avoids researcher bias
• lab experiment

  • iv is manipulated
  • takes place in natural setting
  • set time frame
  • standardised procedure

  😃 high control of variable, high internal validity

  😔 demand characteristic may take effect
• Field experiment

  • manipulate iv
  • takes place in real-life setting

  😔 less control over variables , extraneous variables

  😃 high mundane realism, high ecological validity
• Quasi experiment

  • iv is pre-existing
  • no variable manipulated

  😃 controlled conditions, high validity

  😔 no random allocation of participants
• natural experiment

  • study and occurring situation
  • no manipulation if iv
  • iv is preexisting

  😔 rare so therefor not generalisable

  😃 high external validity, study real life issue that cat be created due to ethical issues
• Pilot study= small scale experiment

  • identify problems/issues with he study
  • make sure it all works
• questionnaires

  😃 collect large amount of data easily and cheap

  😃 do not =need to be administrated by a specialist

  😔 answers not always truthful

  😔 bias sample- only some fill out questionnaires
• Closed questions= have to select from options

  😃 quantitative date, day to analyse

  😔 cannot express feelings, may be to a degree

  open questions= give opinions and feelings

  😃 rich and useful data

  😔 difficult to summarise and detect clear patterns
• structured interview= all questions decided in advance

  semi-structures= some decided inadvane, allow flexibility

  😔 interviewer bias, social desirability

  😃easily repeated

  unstructured= questions lead by the response of participant

  😃 more detailed info can be extracted

  😔 requires well trained people
• nominal data(category data) = only mode can be used

  ordinal data( bankable data) = median and mode

  interval and ratio(on a scale)= mean, mode and median
• measured of dispersion

  • range= difference between the highest and the lowest value in a set of data
  • standard deviation= uses interval/ ratio data

  -provides a measure of how spread scores are around the mean in a set of data

  -large SD= sores are widely distributed around the mean

  -small SD = the scores are closely grouped around the mean

  -help evaluate the value of research studies claiming the relationship between variables

  -large deviation suggests findings ate not trustworthy as it claims to show a relationship between two variables