Research Methods

Sampling

• The process of selecting participants from a population of interest
• Sampling is essential to make sure researchers can minimise extraneous and confounding variables

Random Sampling

• Every member of the population of research interest has an equal chance of being selected to be part of a sample
• This method of sampling helps ensure the sample is highly representative, the larger the sample the more likely this is going to occur
• However this method of sampling can only happen if there is a full list of the target population is available, and this can be hard to obtain

Stratified Sampling

• Dividing the population into different sub-groups (strata) 
• Then selecting a separate sample from each sub-group (stratum) in the same proportions as they occur in the population of interest
• Examples of stratum/subgroups are socio-cultural factors like
  • Residential area
  • Type of accomodation
  • Income level
  • Age
  • Sex etc
• This method of sampling gives the researcher the ability to sample from specific groups within populations for comparison purposes, giving the study greater precision
• However this method of sampling can be time consuming and expensive since it takes up people's time

Convenience (Opportunity) Sampling

• The researcher may go to locations known to be frequented by the required participants 
• The researcher will then select the first individuals they meet who are in the target population and willing to participate 
• The researcher may do this because the target population is not readily available
• Convenience/opportunity sampling is quick, easy and inexpensive
• However this method of sampling often produces a biased sample as researchers may purposely choose participants willing to cooperate and may leave other people and their opinions out of the study
• Results from these samples can't be generalised to the entire population

Independent Groups Design

• Participants only take part in one condition of the experiment (i.e. two different groups exposed to 2 different variables)
• Can all be done at once an drop-outs are unlikely 
• But a large number of participants are needed so the participant variables in the sample can match the population 

Repeated Measures Design 

• Participants take part in both of the conditions of the experiment (i.e. everyone in one group is exposed to two variables) 
• The participants are the experimental and the control group so the researcher must consider any order or practice effects
• Individual participant variables are controlled and fewer participants are required 
• But order and practice effects may do better (they know what to expect during the experiment) or worse (they know it will be boring or they are tired)
• Participants may drop out because this experimental design takes a long time

Matched Particpants Design

• Participants are matched in each condition for characteristics that may affect their performance 
• For example, participants who scored a 3 on a memory test are paired with someone who also got a 3
• The variable which participants have been matched won't affect the results because the researcher has matched people with similar abilities 
• However this is time consuming and expensive as lots of testing is done to find an appropriate pair and if one person from a pair drops out, the remaining person can't continue 

• The way that participants are placed into the groups for the study (these groups are often experimental and control groups)

Random Allocation

• A procedure used to place participants in a group so they are as likely to be in one group as the other

Independent Variable

• The variable that is changed in order to measure its effects on the dependent variable 

Dependent Variable

• Used to measure the effects of the independent variable
• Called the dependent variable because whether it changes or not depends on the effects of the independent variable

Extraneous Variable

• Any variable (other than the independent variable) which may influence the dependent variable 

Confounding Variable 

• Any variable (other than the independent variable) that has influenced tha dependent variable