Research Methods

Research Methods

Aims:

• Hypothesis is a proposed explanation of an observation while an aim identifies the purpose of an experiment.

Hypothesis:

• Directional - a prediction made regarding a positive or negative change.
• Non-directional - predicts that the IV will have an affect on the DV but not how.
• Null - there is no statistical difference between two variables.

Variables:

• Independent variable - the variable that is changed.
• Dependent variable - the variable that is measured.
• Operationalisation of variables - how you will define and measure a specific variable as it is ussed in your study.
• Extraneous variables - undesirable variables that affect the relationship between IV and DV.
• Confounding variables - factors other than the IV that may cause a result.

Sampling:

• The population is the total group of individuals a sample may be taken from but a sample is the group of people who take part in teh experiment to represent the population.
• Random sampling - everyone in the target population has an equal chance of being selected. People are randomly picked from a list. Free from researcher bias. Difficult and time consuming as it can take time to get a list of everyone. It can be unrepresentative and people may refuse to take part.
• Systematic smapling - evry nth member is selected. List of people in target population is formed (sampling frame) and every nth numbered person is pciked. Avoids researcher bias and is usually farily represntative.
• Stratified sampling - target population is divided into sub groups (strota) adn the number of people picked from each strata is proportionla to the number in the whole population. Use percentages. Avoids researcher bias. Is generalisable as it is normally representative. Can't reflect all the ways people are different.
• Opportunity sampleing - select anyone who is willing and available. Asks anyone who is around at the time of their study. Saves time, effort and money. Unrepresentative as all particpants are from one street/tow so can't generalise. Reasearcher has complete control so there is researcher bias.
• Volunteer sampling - participants select themselves. Easy and requires minimal input from the researcher. Biased as it may attract a ceratin profil of person so can't generalise.

Pilot study:

• A small scale preliminary study to evaluate feasablity, time, cost, adverse events and improve the design prior to the full-scale experiment.

Experimental design:

• Repeated measures - using the same subjects for all conditions (IV). Partcipants are controlled and less are needed. The first task could have an effect on the second so counterbalancing is needed. Repating tasks could create boredom/fatigue which could cause performance to deteriorate but could also give practise to improve performance (order effects). Acts as a confounding variable.
• Participants could work out the aim and show demand characteristics.
• Indpendent measures - separate groups of participants for each condtion (IV). Order effects aren't a problem and particpants are less likely to guess the aim. Participants aren't the same and it is less economical as each participant only contributes one result.
• Matched pairs - two groups…