Stages Of A Practical Project
Decide on research methods:
It can be a combination of these, Bandura for example used an observation technique in a laboratory setting.
Decide on type of experiment:
Strengths and Weaknesses of Experiment Types
Strengths: More control of the experiment, more ethical as it easy to get consent.
Weaknesses: Lacks ecological validity-not in a natural setting, high demand characteristics as they can see what is expected of them.
Strengths: Higher ecological validity, lower demand characteristics.
Weaknesses: Can be unethical: difficult to get consent eg: Piliavin, experimenter has less control.
Strengths: High ecological validity as there is no tampering with the IV
Weaknesses: Difficult to repeat as the IV is naturally occuring.
Design of The Experiment
Independant Measures Design:
Strengths: Less demand characteristics as they will not learn conditions
Weaknesses: More difficult to compare as there could be several reasons such as gender etc.
Repeated Measures Design:
Strengths: More accurate and easy to compare the results, costs less than getting twice as many ppts.
Weaknesses: Demand characteristics ppts may become familiar with what is expected.
Strengths: Less demand characteristics as they won't be learnt over time.
Weaknesses: Using different ppts in same conditions not as accurate as originally thought.
Since evidence will either support a prediction or challenge it there has to be two hypotheses:
-A hypothesis that will be accepted if findings support the researcher's expectation-alternate hypothesis
-A hypothesis that will be accepted if the findings do not support the researcher's expectations-null hypothesis
The alternate hypothesis is a testable statement that proposes the expected outcome of the study. In the context of an experiment this means suggesting a significant difference between the results of the two conditions. (The IV will have caused those in the experimental condition to behave in such a radically different way to those in the control that is beyond any difference that could be due to chance.)
The null hypothesis predicts that the findings are not consequences of the predicted effect but due to chance. It enables the researcher to draw a conclusion ie: that any effect there may have been was so small it may have arisen by chance.
Two tailed hypotheses are given this name because they entail a prediction that the outcome could fall in one of two directions. One tailed hypothesis are predictions that the outcome will fall in one direction.
Operationalising a hypothesis
A hypothesis should contain a clear statement of the IV and the DV as well as the word "significant" and an indication as to whether it is one or two tailed. For example:
"People in a village will post on a significantly greater number of apparently mislaid letters than people in a city"
This hypothesis makes it clear what the IV and DV will be, the hypothesis is expecting a result making it one tailed and the hypothesis makes the significant difference clear.
However while this is one tailed the null hypothesis would be non directional:
"There will be no significant difference in the number of apparently mislaid letters posted by people in a village and people in a city: any difference will be due to chance."
The use od the word difference establishes that the experiment has two conditions rathe than a correlation study.
Deciding on the Procedure
You have to decide on the:
IV: If looking for differences between people, you need two groups of participants (matched pairs or independant) if looking at differences between situations you can use independant.
DV: This has an effect on the materials used to measure the DV. Eg: a questionnaire, or an observation) As well as type of data (quantitative, qualititative) and the level of measurement (nominal, ordinal, lateral).
Controls: To ensure that any experiment is valid, controls are needed. They can relate to the ppts (same age, same intelligence), the situation (ensuring data is collected a certain time each day) or the researcher (double blind-person collecting data is unaware of what the experiment is measuring)
Stay within ethical guidelines: consent, deception, withdrawal, avoid causing harm, confidentiality and debriefing.
What participants will actually do- a step by step description of participants should actually do.
Decide on the Sample
Opportunity (using whatever participants are in the area at the time)
Strengths: Varied selection of participants.
Weaknesses: Can be ethnocentric
Random (everyone has equal chance of being selected)
(Stratified is like random but with more theory behind it)
Strengths: Completely fair, mixture of people
Weaknesses: No theory
Self Selected (Volunteers)
Strengths: Ethically sound
Weaknesses: Not always a varied sample.
Analysis Of The Data
Once results are taken researchers must do a statistical analysis of their data and do this through inferential statistics, the test used is determined by two things:
-The design of the experiment.
-The level of measurement of the data that's been collected.
Independant Measures Repeated Measures Matched Pairs
Nominal Chi-Square Sign Test Sign Test
Ordinal Mann-Whitney U Test Wilcoxon Signed Ranks Wilcoxon Signed Ranks
Interval/Ratio Independant t-test Related t-test Related t-test