Self report refers to any method which gets people to report on their thoughts, feelings or behaviour. The main two types of self report are questionnaires and interviews.
Questionnaires can take place over the phone, by email, or as a written questionnaire. They can include rating scales. These are scales that categorise answers or present the participant with a numbered scale. There are different types of questions which you can get on a questionnaire. (1) open questions- these questions allow the participant to answer how they would like to, there are no restrictions. They usually generate qualitative data which is good because it gives you an insight into people's feeling and precision, but bad because it is difficult to categorise or put into descriptive statistics. (2) closed questions- these provide answers for the participants, of maybe simple 'yes' and 'no' options. They restrict the answers participants can give. They usually generate qualitative data. This is good because it is easy to analyse and compare, but bad because it doesn't open up any new avenues and people may not be able to answer exactly how they would like to. (3) leading questions- these are questions which suggest a preferred answer, for example 'was that film boring?'.
Interviews usually take place by telephone or face to face. Nerves can have different structures (1) structured- these interviews have predetermined questions which the researchers stick to. (2) semi structured- these interviews begin with structured questions but can evolve and questions can be asked in response to the participants answers. (3) unstructured- these interviews begin with no pre determined questions and the questions develop as you go.
Things to consider with self report
-Clarity- questions need to be worded clearly. If a participant doesn't understand the question properly then their answer is worthless and will reduce the validity of the data collected.
-Bias- as mentioned previously, leading questions are biased. They push the participant to answer in a certain way.
-Social desirability- participants are always likely to answer in a way which makes them seem to be a better person, they are not necessarily honest. If a question asks 'do you lie a lot' then people are more likely to say no.
-Analysis- the way data is analysed depends on the type of questions asked. Closed questions present quantitative data whereas open questions present qualitative data which represents people's thought and feelings.
Rating scales categorise answer such as 'strongly disagree', 'disagree', 'agree', 'strongly agree' or numerically categorise answers such as, how far do you agree? 1 being strongly disagree and 10 being strongly agree. Respondents are asked to give their answer in the form of a rating.
You would use rating scales in self report methods to get a personal opinion on something but still retain quantitative data.
Experimental method of collecting data
The researcher manipulates an independent variable to see what effect it has on a dependent variable whilst attempting to control all other extraneous variables.
There are different types of experiment:
(1) laboratory experiment- The researcher manipulates independent variables whilst strictly controlling extraneous variables through a standardised procedure in a highly controlled environment.
(2) field experiment- the researcher manipulates the independent variable but the experiment takes place in the participants natural environment.
(3) natural/quasi experiment- the independent variable is naturally occurring, the researcher just records the effect on the dependent variable. Quasi experiments are experiments where control is lacking over the independent variable.
Types of experiment
There are two types of experiment. Independent measures and repeated measures.
-independent measures- independent measures experiment use different groups of participants for each condition.
STRENGTHS- there are no order effects such as boredom, fatigue, practice.
WEAKNESSES- there may be participant variables between the different groups.
SOLUTION- participants could be 'matched' in each group, eg same age, same sex, same IQ, participants could also be randomly allocated to groups.
-repeated measures- the same group of participants are used for all conditions.
STRENGTHS- there are no participant variables.
WEAKNESSES- there may be order effects such as boredom, fatigue, practice.
SOLUTION- counterbalancing, half of one group do the one condition, the other half do the second condition, they then swap.
Observational method of collecting data
Behaviour is observed, recording and then analysed in order to try and make sense of it. Observations involve the precise measurement of naturally occurring behaviour. There are different types of observations.
(1) naturalistic- the observations are made in an environment that hasn't been altered and is as it would be usually.
(2) controlled- the observations are made in an environment where some variables have been controlled, such as a laboratory experiment.
(3) participant- the researcher gets involved with the practices/everyday life of the participant either with or without their knowledge. Can limit objectivity of the researcher, can create researcher bias.
(4) non-participant- the researcher merely watches and records the behaviour of the participants.
Strengths and weaknesses of observation
-what people say and what people actually do are two different things. Observations allow us to capture what people actually do and the difference in that and what they said they'd do.
-observations enable us to capture spontaneous behaviour.
-if it is a participant observation, the researcher can offer special inset into why people have behaved how they have.
-observations allow us to get honest actions/behaviour.
-observations don't allow for any insight into behaviour
-they can arouse ethical issues such as deceit, invasion of privacy etc.
-observation are difficult to replicate
-observers may see what they want to see (observer bias)
-if people know they are being observed then they may alter their behaviour.
TIME SAMPLING: the observer decides in a set time period, for example 2 minutes and observes behaviour of participants for that amount of time. They then note down any behaviour the target individual(s) displayed at the end of the time period.
EVENT SAMPLING: the observer uses a coding scheme and makes a note/tally each time a particular behaviour occurs.
A correlation measures the relationship between two or more variables to try and establish a pattern between them.
Correlations range from 'perfect negative' correlations to 'perfect positive' correlations.
No manipulation of variables is required
Strong significant correlations can suggest ideas for experimental studies to try and establish cause and effect.
They do not show cause and effect.
Some patterns that are found do not yield a significant correlational result.
If two variables increase together, they are a positive correlation. If one variable decreases as another increases, this is a negative correlation.
The hypothesis is made by the experimenter at the start of the experiment and states what they think will happen and what they expect to discover.
A one-tailed hypothesis, such as 'children who listen to music whilst studying will not do as well in exams' states the expected direction of results.
A two-tailed hypothesis Is one where the direction of the hypothesis is not predicted, so, 'children who study whilst listening to music will do differently in an exam than those who don't'.
Both one-tailed and two-tailed hypotheses are alternate hypotheses. They state an expected direction.
A null hypothesis is a statement of no difference or relationship, so, 'there is no difference between performance in an exam by those who listen to music whilst studying and those that dont'. It is 'null'- a statement of nothingness.
Independent and dependent variables
An independent variable is a variable manipulated by the researcher.
The dependent variable is the variable that is effected by the manipulation of the independent variable. It is usually the dependent variable that is measured in an experiment.
Extraneous variables are any other variables apart from the independent variable which might effect the dependent variable. If changes in a dependent variable are due to extraneous variables rather then the independent variables then the results drawn from an experiment may be wrong.
Extraneous variables can be participant variables or situational variables.
PARTICIPANT VARIABLES are characteristics of participants which may influence the outcome of a study, such as, age, intelligence, motivation, experience and gender. Participant variables only really matter in independent measures design where two different groups of participants are used in two different conditions. These participant variables can be controlled by matched pairs, random allocation to groups or by using repeated measure design.
SITUATIONAL VARIABLES are any feature of a research situation which influence a participants behaviour, such as, order effects (practice, boredom, fatigue effects), time of day, temperature, noise, investiator or experimenter bias, demand characteristics.
These situational variables can be controlled by using independent measures design. Environmental and temperature factors can be controlled by keeping them constant. Possibilities to control investigator bias and demand characteristics are single blind, these are experiments in which participants aren't told the real reason behind the study so this discourages them to seek cues to adjust their behaviour accordingly and standardised instructions so each participant receives the same set of instructions and no hints can be given.
EXTRANEOUS VARIABLES ARE PARTICIPANT OR SITUATIONAL VARIABLES-ANYTHING THAT ISN'T THE IV!!
Validiry and reliability
VALIDITY- to what extent has the researcher measured what they wanted to measure
RELIABILITY- how consistent is a measure. If observations or experiments are reliable, then we would expect to come up with the same results if we did them again. Two observers should be able to produce results that are 80% the same, this is inter rater reliability.
Types of sampling
RANDOM SAMPLING- participants are chosen at random. For example, putting the names of everybody in a hat and picking out names.
STRENGTHS random sampling in large numbers provides the best chance of a representative and unbiased sample.
WEAKNESSES the larger the target population, the more difficult it is to sample randomly as getting a list of everybody becomes more difficult.
OPPORTUNITY SAMPLING- involves selected subjects which are around and available at the time. Effort can be made to not be biased in the selecting of participants.
STRENGTHS it is quick, easy and convenient. Probably the most common method of sampling.
WEAKNESSES it is very unrepresentative and can be biased on behalf of the researcher who may chose people who look helpful.
SELF SELECTING SAMPLE- these are participants who have volunteered themselves to take part in the study or experiment.
STRENGTHS self selecting samples are relatively convenient and if volunteering is made on the basis of informed consent, then it is ethical. The choice is not biased by the researcher.
WEAKNESSES generally only a certain type of person volunteers themselves. They are usuall highly motivated, and have a lot of spare time. These samples are often unrepresentative.
-consent- did the participants know what they were letting themselves in for? Have parents given their children consent?
-Deception- have the participants been deceived at all? Was there any other way to carry out the study with out deception?
-debriefing- when the study is over have the participants been told what was happening and asked if they had any concerns? Has the stress caused by any problems been removed?
-observational research- unless consent has been given, observation can only been made in public areas where everyone would expect to be seen by strangers
-right to withdraw
-protection of participants- investigator/researcher must protect participants from mental and physical harm.
-mean- adding up all the numbers and dividing it by the amount of numbers there is.
-median- middle value in an ordered list.
-mode- most common value.
Graphs (used to see results at a glance):
-bar chart- height of the graph represents the frequency, suitable for words and numbers
-pie chart/pictograms- illustrating frequency of data by using slices of pie or pictures.
-scatter graph- suitable for correlation data.
Different kinds of data which can be collected
Nominal- data is in separate categories, eg grouping people by their football team.
Ordinal-data is ordered in some way, eg asking people to put foot ball teams in order of their liking.