case study pros and cons

fr

?
  • Created by: Beth
  • Created on: 07-01-13 23:11

Case Study - +'s and -'s

+ Rich in depth data can be gathered so info that may be overlooked using other methods likely to be identified. Useful for understanding the subtleties and complexities of an individual’s behaviour.

+ Can be used to investigate instances of human behaviour/experiences that rare which perhaps would not be able to generate such conditional experimentally ethically.

+ Data from several people can be pooled and analysed e.g. brain damage patients allowing a greater understanding of causes the symptoms they share.

+ Complex interaction of factors can be studied in contrast with experiments where many V's held constant.

- It’s difficult to generalise from individual cases as each has unique characteristic.

- Use recollection of past events as part of the case history -unreliable.

- R’s may lack objectivity as the get to know the case or because theoretical bias may lead them to overlook aspects of the finding.

- Ethics:confidentiality – cases easily identifiable due to unique characteristics even if real names not given.

- Time consuming and relationship between the R and the individual makes it diff to rely on objectivity of data.

1 of 22

Content Analysis - +'s and -'s

A kind of observational study in which behaviour is observed indirectly in written or verbal material such as (painting, interviews, books, TV.)

It’s indirect (since you are observing people through the artefacts they produce)

+ High ecological validity because it’s based on direct observations of what people actually do; real communications which are current and relevant e.g. recent newspapers.

+ Findings can be replicated because data sources are public/retained.

- Observer bias reduces the objectivity and validity of the findings.

- Likely to be culture-biased because interpretation of content will be affected by language/culture of observer and behavioural categories used.

2 of 22

Processes involved in Content Analysis

The R has to make 2 decisions:

Sampling method: What material to sample and how frequently? (which TV channels, how many programmes and what length of time)

Behavioural categories to be used:

Quantative Analysis: Examples in each category are counted e.g. Content analysis of teen behaviour from letters in a teen magazine e.g. count no. of letters in a teen magazine about categories developed e.g. bulling/sex/health

Qualitative analysis: Examples in each category are described e.g. Content analysis of teen behaviour from letters in a teen magazine e.g. quote from different letters about categories developed e.g. bulling/sex/health

Processes of Content Analysis (Qualitive) - after interviewing P's about family involvement in school 

  • All answers to the same questions put together
  • Each statement developed into a briefer statement and given a code
  • Statements compared with others and categorised  (so statement with similar content placed together
  • Categories grouped into larger units  producing main categories e.g. support/enablement
3 of 22

Aims and Hypothesis

  • Aim:  Intended purpose of investigation
  • Hypothesis: Precise, testable statement written in future tense about target pop.
  • Operationalised Hypothesis: Make it testable, so it can be repeated, increasing reliability of findings. Must operationalise the variables (IV and DV–ensuring they are in a form that can be easily tested) – e.g. how you are going to measure your IV/DV. E.g. 'The scores obtained on a memory test by 10 femalesaged 16-24 will be higher than the scores obtained by  10 males aged 16-24'.
  • Research hypothesis: Proposed at the start of R and is often based on theory
  • Experimental (experiments, H1)/Alternative Hypothesis (observations/opinions, HA): The prediction you are making e.g. evidence there is a significant relationship/ difference between two sets of data.
  • Null hypothesis (Ho): Backup hypothesis; statement of no difference/relationship. If data doesn’t support Ho, reject it and go with HA instead.
  • Directional hypothesis (One-tailed Hypothesis): States expected direction of the predicted difference b/ween two conditions or 3 groups of P’s  e.g. P’s do hmwk without music produce better results than P’s who do hmwk with music. Previous R/pilot study may suggest direction for findings. Easier to reject than a non-directional, so R that proves a directional hypothesis is regarded highly.
  • Non-directional hypothesis (Two-tailed Hypothesis): Predicts that there will be a difference but not the direction of the difference between two conditions or groups of participants ; e.g. P’s who do homework without music will produce different results to P’s who do homework with music. (Use if you don’t know the answer to the problem or think something might happen – next piece of R you may choose directional.)
4 of 22

Variables

 Variable: Something which is observed, measured, controlled or manipulated.

Operationalised Variables: how they will be specifically measured in the study.

Independent variable (IV): Variable the experimenter manipulates – assumed to have direct effect on the DV.

Dependent variable (DV): Variable that is measured, after making changes to the IV.

Extraneous variables (EV’s): Variables other than the IV that may affect dv e.g. temp. If affects all conditions equally confounding/bias does not occur but if EV isn't controlled/can provide alternative explanations for effects = Confounding variables

5 of 22

EV's - Situational Variables 1

Situational variables: Features of a R situation that may have influenced P behaviour and act as EV’S.

They should be controlled to ensure they are the same for all P’s.

Order effects: improved P performance may be due to practise (an EV) rather than the IV.

Time of day, temperature & noise: Only affect the DV if the environmental factor affects performance (e.g. in task is cognitive time of day may be significant as more P’s are alert in the morning) and if it varies systematically with the IV (P’s in group 1 are tested in the morning and group 2 in the afternoon) but if some of each group tested in the morning and others in the afternoon then time of day would not be an EV as it would not have a systematic effect on the DV. 

6 of 22

EV's - Situational Variables 2

Investigator effects: Anything the investigator does which has an effect on P’s performance in a study, other than what was intended (e.g. the investigator’s expectations about a study or the P’s reaction to the behaviour/appearance of an investigator.) E.g. the way the investigator responds to a P may encourage some P’s more than others e.g. male R’s more encouraging with female P’s .  Includes - >

  • Direct Effects: (consequence of the investigator interacting with the P) and indirect effects (consequence of the investigator designing the study.) Effects of this are greater in non-experimental investigations such as interviews than observations.
  • Indirect Effects:Investigator experimental design effects: Investigator may operationalise the measurement variables in such a way that the desired result is more likely or may limit the duration of the study for the same reason.
  • Investigator loose procedure effect: Investigator may not clearly specify the standardised instructions or/and procedures leaving room for the results to be influenced by the experimenter. These should be controlled to ensuring they’re the same for all P’s.
  • Demand Characteristics: Cue makes P’s aware of what the R expects to find/how P’s are expected to behave. Can change the outcome of a study because P’s change behaviour to conform to expectations.
  • Screw You Effect  P knows what is happening in the exp  and purposely refrains from showing any interest in it.
  • Please You Effect – When a P knows what is happening in an exp and tries to alter results to please the experimenter even though realistically they won’t act that way.
7 of 22

Controlling Situational V's and P effects

Standardised Procedures: This means that each P is treated in exactly the same way, each doing exactly the same tasks, with the same materials, in exactly the same order. This reduces the variables in the procedure.

Standardised Procedures include Standardised instructions: Each P must be given exactly the same instructions, ideally by the same person and in the same way; otherwise this could affect the results. You can ensure standardisation by providing written instructions, which are simple and clear.

Double blind design: Neither the P’s nor experimenter is aware of the aims/important details of the study so have no expectations.

Controlling participant effects:

Single blind design: When P’s do not know the true aims of the study so cannot seek cues about the aims and react to them. 

Experimental realism: The extent to which P’s become involved in an experiment and become less influenced by cues about how to behave e.g. by making the experimental task more engaging, P’s are less likely going to be looking for cues for how to behave.

8 of 22

Demand Characteristics and Investigator Effects

Demand Characteristics- A cue that makes P's aware of what the R expects to find or how P's are expected to behave. These can change the outcome of a study because P's will chang their behvaiour to conform to expectations.

Screw You Effect – P knows what is happening in the experiment and purposely refrains from showing any interest in it.

Please You Effect – When a P knows what is happening in an experiment and tries to alter results to please the experimenter even though realistically they won’t act that way.

Investigator Effects - Anything the investigator does which has an effect on P’s performance/outcome of study, other than what was intended (e.g. the investigator expectations about a study or the P’s reaction to the behaviour/appearance of an investigator. Includes both direct (consequence of the investigator interacting with the P) and indirect effects (consequence of the investigator designing the study.) Effects of this are greater in non-experimental investigations such as interviews than observations.

Indirect Effects: Investigator experimental design effects: Investigator may operationalise the measurement variables in such a way that the desired result is more likely or may limit the duration of the study for the same reason.

Investigator loose procedure effect: Investigator may not clearly specify the standardised instructions or/and procedures leaving room for the results to be influenced by the experimenter. These should be controlled to ensuring they’re the same for all P’s.

9 of 22

What is Experimental Design?

A set of procedures used to control the influence of factors such as P variables in a experiment.

The 3 experimental Designs are -

Independent Groups

Repeated Measures 

Matched Pairs

10 of 22

Experimental Design - Independent Groups Design

2 or more separate groups. P’s randomly allocated to one of the conditions. Each group tested in a different condition (1 of them being controlled). 

+ Avoids order effects

- Potential for error resulting from individual differences/P variables between the groups of P’s taking part in the different conditions. 

-More time consuming/expensive - Twice as many P’s are needed than with the repeated measures design.

How can these limitations be overcome

 -P’s variables can be overcome if the sample size is large enough and if P’s are randomly allocated (theoretically distributes P variables evenly.) Random Allocation - Allocating P's to experimntal groups or conditions using random techniques

- Spend more time and money

11 of 22

Experimental Design - Repeated Measures

Each P takes part in every condition under test.

+ Individual differences/P variables between P’s are removed as a potential cofounding variable.

+ Fewer P’s required, since data for all conditions are collected from same  P’s which is quicker and cheaper.

- Range of potential uses is smaller than for independent groups design E.g. reading schemes 

- 1 condition may be harder than another (EV) - affect accuracy of results.

- On 2nd test, P’s may have guessed experimental aims and may influence answers - P effect

- Order effect: EV arising from the order in which conditions are presented. May affect performance through getting better through practice (learning effect) or getting worse through being boredom/tired (fatigue effect). 

How can these limitations be overcome

  • Make equivalent tests to make both conditions equal.
  • Order effects can be controlled by counterbalancing
  • Use single blind - A type of R design if which the P is not aware of the R aims or of ehoch conditions of the experiment they are recieving
12 of 22

Counterbalancing

Order effects can be controlled by counterbalancing: alternating the order in which P’s perform in different conditions of an experiment to balance the effects across both conditions or by having a sufficient time delay between 2 conditions.

Ensures that each condition is tested first or second In equal amounts, AB/BA or ABBA.

A- morning test           B- afternoon test

  • AB/BA
  • E.g. Group 1 each P does 'A' then 'B'
  • Group 2 each P does  'B' then 'A'   - Comparision made for each P on their performance on the two conditions (morning and afternoon)
  • ABBA - All P's take part in each condition twice
  • Trial 1 - Condition  A (moring)
  • Trial 2 - Condition  B (afternoon)
  • Trial 3 - Condition  B (afternoon)
  • Trial 4 - Condition  A (moring) - Compare scores on trial 1+4 with 2+3, still rep. measures as comparing the scores of the same Person
13 of 22

Experimental Design - Matched Pairs Design

Pairs of P’s are matched in terms of key variables such as age/IQ. 1 member placed in the experimental group, the other in the control.

+ No order effects as P’s are only completing one condition.

+ Lowers P variables (although still no full control) produces more valid results. 

-Achieving matched pairs of P’s is  difficult/time consuming task and may be too costly as you must start with lots of P’s to ensure you can obtain matched pairs on key variables. 

- Impossible to match people exactly, unless identical twins

- May not control all P variables as you can only match on variables known to be relevant, but others could be important.

How can these be overcome

  • Restrict matching variables to make it easier.
  • Use identical twin pairs; provides a good match.
  • Conduct pilot study to consider key variables.
14 of 22

Experimental/Control Groups or Conditions

Independent Groups - have experimental and control groups as (each P is assigned to 1 group)

Repeated Measures - have experimental and control conditions (each P experiences both conditions)

15 of 22

Ethics

Ethical Issues: A conflict in what the R needs in order to conduct useful and meaningful R and the rights of P’s. Ethical issues are conflicts about what is acceptable.

Ethical Guidelines: Concrete, quasi-legal documents help to guide conduct by establishing principles for standard practice/competence (way of resolving the conflict)

Theses ethical principles are set by the BPS and failure to follow them can lead to psychologists being rejected from the society or their licences being revoked and their name and R blackened.

What are the ethical Issues?

  • Informed Consent
  • Deception
  • The Right to Withdraw
  • Protection from Hram
  • Confidentiality
  • Privacy - we have a right to privacy if this is invaded than confidentaility should be respected
16 of 22

Ethical Issues - Informed Consent

P’s have the right to be given comprehensive info concerning the nature/purpose of the Rand their role in it, in order that the can make an informed consent about whether to participate and assess the risk factor.

How to deal with it:

  •  P’s asked to formally indicate their agreement to participate and this should be based on comprehensive info concerning nature/purpose of the R and their role in it. 
  • Alternative is to gain presumptive consent (gain consent from others rather than P’s.)
  • R’s can also offer the right to withdraw
  • If anyone below legal age of consent is used in the R, then consent sought from the parents/guardians, same may apply for P’s with mental illness/learning difficulties/ old dementia patients they’re vulnerable so R must proceed very carefully. 

Limitations:

  • If a P is given info concerning the nature/purpose of study then this may it.
  • Even if R’s have sought and obtained informed consent, this doesn’t guarantee that P’s really do understand what they have let themselves in for.
  • Problem with presumptive consent is that what people expect they will/will not mind can be different from actually experiencing it.
17 of 22

Ethical Issues - Deception

Where a P is not told true aims of the study (e.g. what participation will involve) and thus cannot give truly informed consent so great care and careful consideration must be given to the project/use of one-way mirrors etc.

How to deal with it:

  • The need for deception should be approved by an ethics committee, weighing up the benefits (of the study) against costs (to P’s): cost benefit analysis. 
  • P’s should be debriefed i.e. told they have been lied to (deceived) for the need of an experiment and offered the opportunity to withhold their data.

Limitations:

  • Cost-benefit decisions are flawed as they involve subjective judgements, and the costs aren’t always apparent until after.
  • Debriefing can’t turn the clock back: P may still feel embarrassed/have lowered self-esteem/emotionally distressed for being lied to and if P’s are going to leave in a state different to when they entered they should not be involved.
18 of 22

Ethical Issues - The Right to Withdraw

How to deal with it:

P’s should have right to withdraw at any time regardless of payments if they’re uncomfortable and should have the right to refuse permission for the use of their data.

Limitations: 

P’s may feel they shouldn’t withdraw as it will spoil the study.

In many studies P’s are paid/rewarded so may not feel able to withdraw.

19 of 22

Ethical Issues - Protection from Harm

R’s have responsibility to protect P’s from physical/mental harm. Normally, risk of harm should be no greater than encountered in everyday life. This includes confidentiality.

How to deal with it:

Avoid any risks greater than everyday life

Stop the study.

Limitations

R’s are not always able to accurately predict the risks of taking part in a study. 

20 of 22

Ethical Issues - Confidentiality

Is a legal right failure to keep details confidential means they have failed to fully protect P from harm.

How to deal with it:

R’s should not record the names of any P’s; they should use numbers or fake names.

Limitations:

Sometimes possible to work out who P’s were on the basis of the info provided. So, in practice, confidentiality not be possible.

21 of 22

Ethical Issues - Privacy

Is the ability of an individual or group to keep their lives and personal affairs out of public view, or to control the flow of information about themselves

How to deal with it

Don’t observe anyone without their informed consent unless it is in a public place.

P’s may be asked to give their retrospective consent or withhold their data.

Limitations

No universal agreement about what constitutes a public place.

Not everyone may feel this is acceptable e.g.  lovers on a park bench.

22 of 22

Comments

No comments have yet been made

Similar Ancient History resources:

See all Ancient History resources »