Вы находитесь на странице: 1из 15

Observation

Observational techniques in general Strengths: What people say they do is often different from what they actually do, so observations give a different take on behaviour A means of conducting preliminary investigations in a new area of research, to produce hypotheses for future investigations Able to capture spontaneous and unexpected behaviour

Weaknesses: Observer may see what they expect to see (observer bias) Poorly designed behaviour checklist reduces reliability (low inter-rater reliability) If participants dont know they are being observed there are ethical problems such as deception and invasion of privacy. If participants do know they are being observed they may alter they behaviour Observations cannot provide information about what people feel

Improving inter-rater reliability: Reliability can be improved by making sure that observers are trained in the use of the behavioural checklist. It might also be necessary to review the checklist and see if some categories are unclear or need sub-dividing to make for more accurate coding.

Structured/Unstructured Structured - observer records certain categories of behaviour using a coding system or behaviour checklist. Strengths: Control of confounding variables Replicable and so reliable/Good internal reliability

Weaknesses: Observer effect/bias Participant reactivity Less ecological validity

Unstructured observer records all the behaviours with no preditermined ideas or categories.

Participant/Non-participant Participant the observer may participate in the experiment itself (a type of action research) Strengths: Acceptance from participant group Likely to provide special insights into behaviour, from the inside

Weaknesses: Objectivity may be reduced Observer/effect bias

Non-participant the observer is merely watching the behaviour of the others Strengths: Allows for greater objectivity than participant observation

Weaknesses: Observer/effect bias

Indirect/Direct Direct when the participants themselves are observed Indirect when observations are made of data that has already been collected (eg. TV or newspaper advertisments)

Controlled/Naturalistic (Uncontrolled) Controlled observations occur in laboratories Naturalistic (Uncontrolled) observations occur in natural environments Strengths: High ecological validity especially where people are unaware of being observed Can produce extremely rich data which is particularly useful in revealing origins of abnormal behaviour Low demand characteristics where people are unaware of being observed Can be used to generate hypotheses for future experiments

Weaknesses: Ethical issues, such as privacy and lack of consent. Lack of control leads to confounding variables.

Problems of demand characteristics if people are observed with their permission Difficult to conclude cause-and-effect relationships as no variables are being manipulated Extremely difficult to replicate Can be difficut to ensure inter-rater reliability

Event-sampling/Time-sampling Event-sampling continuous; involves observations of an event each time it occurs throughout the observation period Strengths: Useful when behaviour to-be-recorded only happens occasionally and might be missed if time sampling used

Weaknesses: Observer may miss some observations if too many things happen at once

Time-sampling discontinuous; involves observations at set intervals, eg. for set length of time or making spot observations every minute Strenghts: Reduces the number of observations by using an objective means of sampling

Weaknesses: Observations may not be representative

Disclosed/Undisclosed Disclosed informing the perticipants that they are being observed (may alter their behaviour) Undisclosed observing participants without telling them first, eg. by using one-way mirrors (without informed consent debriefing (explain afterwards the aims of procedures of the study and asking permission to use their data) becomes especially important)

General Study Design


How to design an experiment: 1. 2. 3. 4. 5. Identify and operationalise DV and IV Select experimental design Write the hypothesis Decide on the procedures Choose sampling method

Operationalising and measuring variables Independent variable what is directly manipulated by the experimenter in order to test its effect on another variable Dependent variable what is measured during the study Operationalisating identify what I am measuring (DV)

Hypotheses Hypothesis a proposed explanation for an observed phenomenon. Research hypothesis a general hypothesis (that may not necessarily be tested because of too many variables) Experimental (=alternative) hypothesis a specific hypothesis that can be experimentally tested Null hypothesis a statement of no significant effect (difference or correlation) of one variable or another (eg. there is no effect between A and B) One-tailed (directional) hypothesis predicts the direction of the result for the hypothesis being tested (eg. girls are more intelligent then boys) Two-tailed (non-directional) hypothesis makes no prediction of direction (eg. girls different intelligence to boys)

Sampling methods Random sample selected in a way that means that everyone in the target population has an equal opportunity to be chosen. Often the selection is done on a computer using random number generators, and giving everyone in the target population a number.

Strengths: Gives the best chance of an unbiased representative sample

Weaknesses: It can be more complex and time-consuming than other methods, especially in large target populations when a selection list of everyone becomes impractical

Stratified sample like a random sample, but where certain subgroups are identified and each subgroup is filled at random to achieve the finel total sample size Strengths: More representative

Weaknesses: More time-consuming and rare

Opportunity sample the researcher simply uses people who are available at a particular time or in a particular place (first 5) Strengths: Quick and easy

Weaknesses: Unlikely to be representative of a target population Can be very biased on the part of the researcher who may choose participants who will be helpful

Quota sample like a stratified sample, but where the subgroups are filled by opportnity sampling (not randomly) Volunteer (self-selected) sample one in which people volunteer to take part Strengths: Relatively easy way of achieving a sample Its possible to target/request participants who possess a particular feature

Weaknesses: Unlikely to be representative of a larger population May be biased on the part of the subject who is willing to participate

Systematic sample in which ever 4th, 10th, etc. name is taken from list of the target population, or person walking down the street, etc Snowball sample allowing one participant to lead the researcher to the next participant

Experiment
Field/Laboratory Field performed in a natural environment, where the experimenter manipulates the independent variable Strengths: Greater ecological validity, because surroundings are natural Less likelihood of demand characteristics (if subjects are unaware of the research taking place) Less likelihood of bias from sampling

Weaknesses: Less control, therefore more chance of extraneous variables Difficult to replicate Difficult to record data Time-consuming Ethical problems of consent, deception, invasion of privacy, etc

Laboratory occurs under controlled conditions Strengths: Increased control of variables Manipulation of independent variables can indicate a cause-and-effect relationship Standardised procedures mean that replication is possible

Weaknesses: Low ecological validity can cause unnatural behaviour Low experimental realism (participants forgetting that in an experiment) Likelihood of low mundane realism (extent that the action is normal) Likelihood of sampling bias May be demand characteristics May be experimenter bias May be ethical issues May be expensive Total control of variables never possible

Independent/Repeated/Matched Pairs Experimental Design Independent Measures Design different participants experience each condition. Each group of participants will only experience one condition

Strengths: No order effects Same stimulus material Drop-out less likely Demand characteristics reduced

Weaknesses: Greater participant variables this may affect the results More participants required

Repeated Measures Design the same participant experiences each condition. Each group of participants will experience both conditions Strengths: Reduced participant variables Participants are the same in each group, so avoid individual differences Less participants required

Weaknesses: Order effects (participants become more skilled as they do the same task twice) May need different stimulus material (different tasks introducing a further variable) Drop-out more likely (attrition) Demand characterictics increased

Matched Pairs Design a variation of the independent groups design, where different participants, who are closely matched (by various factors such as IQ score, gender, age, personality, family background, etc) experience each condition. Therefore, each half of a pair of participants will only experience one condition Strengths: No order effects Same stimulus material Drop-out less likely Demand characteristics reduced Reduced participant (extraneous) variables

Weaknesses: Difficult to match participants exactly Long and time-consuming process of pairing that can be quite wasteful of participants as many have to be tested to find mathing pairs

True/Natural True where the experimenter can randomly allocate participants into one group or another Natural where the experimenter cannot randomly allocate participants into one group or another (where the independent variable changes naturally) Strengths: Allows researchers of areas that would otherwise be unavailable to them Increased validity of findings due to lack of experimenter manipulation Very little participant or situational bias

Weaknesses: Difficult to inter cause-and-effect due to lack of control and since no manipulation of independent variable Impossible to replicate exactly Ethical problems of consent, deception, invasion of privacy etc

Self-report
Self-report techniques in general Strengths: Direct reports rather than via observation of their behaviour Large amounts of data gathered quickly and cheaply, so larger sample sizes are possible Replicable Easy to score/analyse

Weaknesses: People may easily be affected by demand characteristics Scales/questions may be interpreted diferently Closed questions can force answers that dont reflect true opinion Open questions can be very diffiult to analyse Can be lacking in richness/detail

Improving reliability Reliability can be improved in the case of interviews, by training interviewers. Improving validity If the scores are not similar then the test should be revised by changing some of the questions.

Interviews/Questionnaires
INTERVIEWS can be Structured/Semi-structured/Unstructured Structured contain fixed pre-determined questions Strengths: Can be easily repeated Easier to analyse than unstructured interviews because answers more predictable

Weaknesses: Interviewer bias (the interviewers expectations may influence the answers the interviewee gives) Reliability may be affected by the interviewers behaviour

Semi-structured guidelines for questions, but phrasing and timing are left up to the interviewer, and there may be more open questions

Unstructured contain a topic area for discussion but little more (eg. what do you think of an ice-cream

Strengths (semi-structured or unstructured): Generally more detailed information can be obtained from each respondent than in a structured interview Can access information that may not be revealed by predetermined questions

Weaknesses (semi-structured or unstructured): More affected by interviewer bias, because interviewer is developing questions on the spot which gives scope for them to ask leading questions Requires well-trained interviewers, which makes it more expensive to produce reliable interviews when compared to structured interviews

QUESTIONNAIRES can be Psychometric/Opinion Psychometric tests eg. personality and IQ tests. Items need to be standartised for a population and tested to show reliability and validity. Opinion surveys eg. attitude scales (Likert scales) and opinion polls. Questions can be open or closed. Strengths of questionnaires: Can be easily repeated so that data can be collected from large numbers of people relatively cheaply and quickly Respondents may feel more willing to reveal personal/confidential information than in an interview

Weaknesses of questionnaires: Answers may not be truthful (because of leading questions) The sample may be biased, because only certain kinds of people fill in questionnaires (who are willing to spend time filling them)

QUESTIONS can be Open/Closed Open questions allowing participants to give a response in his or her own words, and mean that unexpected information can be collected Strengths: They can produce unexpected information Allow correspondents to express what they actually think rather than being restricted by preconceived categories

Weaknesses: More difficult to analyse qualitative data produced and thus more difficult to draw conclusions

Closed questions that require the participants to choose from a range of pre-determined answers (eg. yes, no, maybe etc) Strengths: Easier to analyse data

Weaknesses: May not permit people to express their precise feelings

Likert/Tick-box Likert scale eg. on a scale from 1 to 10, with 1 being not at all and 10 being very much: Do you like ice-cream? Strengths: Enables respondentsto represent thoughts and feelings quantitatively

Weaknesses: Respondents may avoid using ends of scales and go for middle of the road, thus answers do not represent true feelings

Correlation
Correlation shows a relationship between two variables. Results are generally plotted on a scatter graph that displays the direction and strength of a relationship. Positive correlation two variables increase or decrease together. This does not mean that one variable causes the change in the other (there may be a third variable that affects both of these) Negative correlation as one variable increases the other decreases, or vice versa. Again, this does not show a cause-and-effect relationship but simply that the variables are negatively related. Strengths of correlation: Can be used when it would be unethical or impractical to conduct an experiment If correlation is significant then further investigation is justified If correlation is not significant then you can rule out a causal relationship Gives precise information on the degree of relationship between variables

Weaknesses of correlation: No cause-and-effect can be inferred Extraneous variables can lead to false conclusions Technique is subject to any problems associated with the method used to collect data

Evaluation Alphabet
Application to real life how useful is this research to real life situations? Bias the distortion of the results of the study in a particular direction Systematic bias affects the whole experiment Random bias is unpredictable Experimenter bias leading questions demand characteristics

Procrustean bias influence participants performance or behavior Participant bias arising from the participants Situational bias arises from same aspects of the environment Sample bias the way you sample Confounding variables affect the dependent variable Controlled variables kept the same during an experiment (extraneous variable) Data rich detailed info from extensive interviewers (eg. case study) Demand characteristics when participants form an interpretation of the experiments purpose and change their behaviour accordingly The faithful participant role the participant follows the instructions given provided in the experiment to the letter (no bias) The good participant role the participant attempts to discern the experimenters hypotheses and to confirm them The negative participant role the participant attempts to discern the experimenters hypotheses, but only in order to destroy the credibility of the study The apprehensive participant role the participant is so concerned about evaluation of their answers by the expermenter that the participant behaves in a way that shows high social desirability Ethics the extent to which the procedure conforms to a particular cultures values Informed consent/assent agreement, it must be given for something to happen (avoiding undue pressure or excessive payment) Deception propaganda of beliefs that are not true or not the whole true (not telling them exactly what the experiment is looking at) Debriefing telling later about the purposes of the study (not in the beginning) affects the results

Right to withdrawal a participant has rights to drop-out from the experiment when they want to do it Confidentiality some types of communication between participant and a psychologist, that may not be discussed and is kept private Longitudinal vs. Snapshot vs. Cross-cultural studied the same individual over a period of time Mean arithmetic average of a collection of numbers as computed by adding them up and dividing by their number Median the number in the middle of the range (put the values in order from the lowest to highest, then find the number that is exactly in the middle or between the two if there are even number of values) Mode the value that occurs the most often Order effects in a repeated measures design this is a confounding variable that arises from the order in which the conditions are present. E.g. a practice effect, boredom effect or fatigue effect. One way of reducing the chances of an order effect is to use counterbalancing (each condition is tested first or second (or more) in equal amounts) Quantitative data involves numbers Strengths: Allow identification of patterns of behaviour Can calculate averages Can show dispersion/degree of scatter of data Results easy to analyse Allows comparisons between groups/individuals

Weaknesses: Gives no explanation for why the behaviour occurred Looses richness Is fairly superficial

Qualitative data involves words or descriptions Reliability Refers to the consistency of a technique and how easily can be replicated (how likely is it to give the same answer every time when measuring the same thing Internal reliability every time you do the same study and you get the same results (how well standartised) External reliability other people do the same study at a different time after you and get the same results

Inter-rater reliability whether the different observers ratings or codings would agree if they were all observing the same behaviour Validity Internal validity whether a method of measurement actually measures what it is supposed to (whether the outcome of a study is not being distorted by one or more extraneous factors) 1) Face validity 2) Construct validity 3) Concurrent internal validity External validity whether the results can be generalised if conducted in different environments or using different participants (how closely was to the measure that you wanted to measure) 1) 2) 3) 4) 5) Ecological validity Population validity Historical validity Predictive validity Concurrent external validity

Вам также может понравиться