AQA Psychology: Research Methods (Year2)

• Objectivity: An approach to events that is characterised by freedom from interpreative bias or prejudice. Based on assumption that there is an external realitiy not effects by internal interpretation 
• Replicability: That good research should be able to be repeated and similar results achieved
• Theory Constructuion: Facts alone are meaningless - facts/ideas need to be constructed in order for them to make sense
• Hypothesis Testing: Checking whether theories are correct - done by forming a testable statement that can be tested experimentally
• Empirical Methods: Knowledge is gained through observation and the collection of facts through experimentation
• Paradigm: A general theory which is generally accepted by psychologys - these are not fixed and will change if a new theory is developed with better evidence

Stages of Science

• Stage 1: Pre-science - Variety of theories that attempt to explain and event, but there is no generally accepted theory
• Stage 2: Normal Science - A paradigm has emerged that has generally accepted core assumptions - but will change with new evidence
• Stage 3 - The accumilation of evidence against the existing paradigm is such that it is questionable and new theories are deveoped

Peer review is the assessment of scientific work by others who are experts in the same field. The intention or peer reviewing is to ensure that any research conducted and published is of high quality. 

Three main purposes

• Allocation of research funding - Goverment and charitiable bodie who fund research need to decide if its worthwhile
• Publication of research in scientific journals and books - correct incorrect or faulty data 
• Assessing the research rating of university departments - assessing in terms of quality

Evaluation

• Unacheiveable Idea: not always possible to find an appropriate expert with the same specialist interests to review research
• Anoymity: allows a reviews to be honest
• Publication bias: favour the publication of positive results
• Preserves the status quo: a perfecerence for research that doesn't change existing theory
• Already published research: once research is published it continues to be used even if found to be fraudulent
• An alternative: Online blogs and journals invite comments from any reader as a means of paper reviewing

Types Of Experiements 

Lab

• IV is manipulated by an experiementer to observe its effect of the DV. These are highly controlled
• Can draw casual conclusions, Extraneous variables are minimised and Can be easily replicated
• Contrived: tends to lack mundane realism, Investigator bias and Participants effects e.g. demand characteristics

Field

• More nautral surroundings, IV is directly manipulated by experiementer. These are less controlled
• Can draw casual conclusion, Higher ecological validity and Reduced experimenter effects
• Less control, May have demand characteritics and Difficult to control extraneous variables

Natural

• IV is not directly manipulated and participants not randomly allocated
• Allow research where the IV can't be manipulated for ethical/practical reasons and enables psychologists to study 'real problems'
• Cant demonstrate causal relationships, May extraneous variables and Investigator/Participant effects

Repeated Measures - Same participants are used in both condition

• Less participants are needed
• Order effects may occur and tiredness affecting results

Independent Groups - Participants are randommly allocated to different groups which represent the different conditions

• No order effects will occur
• More participants are needed and Individual difference between participants

Matched Pairs - Pairs of participants are closely matched and are then randomly allocated to one of the experiemental conditions

• Invidiual differences are taken into account and No order effects will occur
• More participants are needed

Observational

• Naturalistic: Everything is left as normal and all variables are free to vary - Can study Behaviour where it is possible to manipulate variables and High ecological validity. Poor control of extraneous varibables and Low inter-observer relaibility
• Controlled: Some variables are controlled by the researcher - Can manipulate variables to observe effects. Less natural and Reduced ecological validity
• Content Analysis: Indirect observation of behaviour based on written/verbal matieral - High ecological validity and can easily be replicated. Observer bias

Self-Report

• Questionnaires: set a written questions - Easily repeated because lots of people can be questioned and Respondents amy be more willing to reveal personal info. Leading questions, Social desirability and Biased Samples
• Interviews: Unstructured questions based on respondents answer - More detailed information and Can access unexpected info. Time consuming and Social-desirability bias

Correlational 

• Co-variables: examind for positive, negative and zero correlation- Can be used when it isnt possible to manipulate vairbales. People often misinterpret correlations

Case Studies: Detailed study of a single individual - Rich, in depth data collects and unsed to investigate unsual instances of behaviour. Cant generalise of public

Reliability

• Internal: is a measure of the extent to which something is consistent within itself 
• External: is a measure of consistency over several different occasions

Validity

• Internal: concerns what goes on inside a study - whether the research did test what they intended to test
• External: concerns things outside a study - extent to which the results of a study can be generalised to other situations and people

Reliability

• Reliability refers to the ability to repeat a study and obtain the same results
• Replication are concluded to test the reliability/validity or the orignal results - if the same results are obtained the second time this shows that the results are legitimate
• Important that all conditions are the same when conducting  a replication

Validity

• Internal: affected by extraneous variables at that act as an alternative IV
• External: affected by the contrived and artifical nature of lab experiements

Reliability

• Observers should be consistent which means that ideally two observers should produce the same results
• Assessing reliability: inter-rater reliability is the measure of the extent to which 2 or more observers agree - caluclated by dividing total agreements by total number of observations (+80)
• Reliability can be improved through training observers in the use of e.g. coding systems

Validity

• Observations will not be valid if the coding behaviour checklist is flawed
• Internal: affected by observer bias 
• Observation studies are likely to have high ecological validity because they involve more natural behaviours

Reliability

• Internal: assessed using the split-half method: compare a person's performance on two halves of a questionnaire or test, there should be a close correlations in the score from both halves of the test
• External: assessed using the test-retest method - a person is given a questionnaire on one occasion and then this is repeated again after a reasonable interval, if the measure is reliabile the outcome should be the same

Validity

• Face Validity: does the test look like it is measuring what the researcher intended to measure?
• Concurrent Validity: this can be established by comparing performance on a new questionnaire/ test with a previously established test on the same topic
• External Validty: likely to be affected bu biased sampling strategies. 

Aim: to select a representitive sample from a target population in order to be able to generalise from the sample to the target population. A sample that is not representative is described as biased. A biased sample means that the generalisations lack external validity

Volunteer - Where the sample is attained through advertising and individuals actively choose and contact the researcher to take part

• No researcher bias as they have no idea who will volunteer
• Individual differences: aquire the same type of person who is confident and outgoing

Opportunity - The sample is taken at a place that the researcher is at and they would just take who is avaliable to them

• Requires little effort meaning researcher can obtain a mass sample quickly
• You cant generalise findings because it is an unrepresentative of the target population

Random - participants are selected from a target populaition using a random number technique

• There is no researcher bias because all people have equal chance of being chosen
• bias sample - in terms of gender and age 

Stratified - Involes classifiying the population into categories and then choosing a sample from each category 

• More respresentative than other methods becuase there is propotional representation of subgroups
• opporunitiy sampling may lead to bias

Snowball - Start with one or two people and they pass on the experiement 

• Useful when conducting research with participants who arn't easy to identify
• Prone to bias because only a limited section of the population is contacted

Ethical issues 

• Informed concent: particpants must be given comprehensive information concering the nature and puropise of a study and their role in it
• Deception: This occurs when particpants are not told the true aims of a study
• Right to withdraw: Particpants should have the right to withdraw from the study at any time
• Protection from harm: Particpants should not experience negative physical effects or negative psychological effects
• Privacy: A persons right to control the flow of information about themselves

Dealing with them

• Debriefing: A post-research interview to inform particpants about the true nature of a study
• Ethical Committee: A group within a research institution must apporve a study before it can be conducted
• Ethical Guidelines: Concrete, quasi-legal documents that establish principles for standard practice and competence
• Competence: Psychologists should maintain high standard in their professional work
• Responsibility: Psycholigsts ahve responsibility to clients, the genral public and science of psychology.
• Integrity: Psycholgists should be honest and accurate. This isclude reporting research findings accurately

The Use of Animals

• Animals may be studied becuase they are interested in thier own right and such research may benefit animals.
• Animals offer the opportunity for creater control and objectivty in research prodcedures.
• Animals may be used when we can't use humans e.g. research on the effect of emotional deprivation or effect of drugs to treat mental illness

Sentient beings: Animals do respond to pain but there is little evidence that animals other than primates have self-awareness. However, equally some humans lack sentience

Speciesism: Discrimination on the basis of species is no different to racism and sexism. However, Gray argued that we have a sepcial duty to care to humans so speciesim is not equivalent e.g. Racism

Animal Rights: Singer's view is utilitarian one i.e. if animal research can alleviate pain and suffering it is justifiable. Regan argues that there are no circumstances under which animal research is acceptable

• Significance: A statisitical terms indicating that a set of research findings is suffiecntly strong for us to accept the research hypothesis under test
• Null hypothesis: there is no difference between men and women in term of their ability to read maps
• Alternative Hypothesis: Women are better than men in terms of thier ability to read maps

Explanations For This Difference

• This pattern could have arisen by chance: it would be incorrect to conclude that there is a real difference between the groups. We should accept Null-Hypothesis 
• This pattern did not arise by chance and is described as significant: We conclude that there is a real difference between the gorups. We reject Null and accept Alternative

Significance Level

• The level of probability at which it has been agreed to reject to null hypothesis
• If the results are less than 5% then it is said that the result are significant, that means the results that occured from the experimental hypothesis is 95%

Probability

• A numerical measure of the likelihood or chance that certain events will occur
• p<0,05 - there is less or equal to 5% possiblitiy that the results did occur by chance.
• Obtaining this result if the null hypothesis was true is less than or eqaual to 5%

Type 1 and Type 2 Errors

• Type 1: Rejecting a null hypothesis that is true. This is more likely to happen if the significance level is too high
• Type 2: Accepting a null hypothesis that is in fact not true. This is more likely to happen if the significance level is too low

Inferential Anaylsis

Inferential Tests

• Drawing local conclusions
• Whether a difference is or isnt significant 
• Arisen by chance or whether the effect occured becasue there is a real difference/correlation

Different Inferential Tests

• Spearman's Rho: Correaltion, Data is Ordinal or Interval
• Chi-Squred: Difference, Norminal Data only
• Mann-Whitney U: Difference, Independent groups and Ordinal/Interval Data
• Wilcoxon T: Difference, Repeated Measures and Ordinal/Interval Data

Overseved and Critical Value

• Two values used to determine whether the results of a study occured by chance or are significant
• The value that is caluclated for any set of data is called the observed value
• To desicde if the observed value is significant it is compared to the critical value, found in a table or critical values

Finding the Appropriate Critical Value

Degrees of Freedom: In most cases we get this value by looking at the number of particpants in the study (N)

One or Two-tailed Test: Directional = One Tailed, Non-Directional = Two-Tailed

Significance Level: The research decides on a suitable probability, usually p<0.05

Greater of Less than: Whether the observed value needs to be greater than or less than the critical value for significance to be shown. This information is stated with each table of critical values

Descriptive Statistics - Measure of Central Tendency and Dispersion

Measures of Central Tendency

• Mean: Calculated by adding up all score and dividing by the number of scores
• Median: The middle value as in a order list
• Mode: The value that is more common in a data set

Measures of Dispersion

• Range: Caluclated by finding the difference between highest and lowest score in a data set
• Standard Deviation: Epxresses the spread of data around the mean - more precise measure because of all the values of the data are taken into account. However, some characteritcs of the data are not expresssed such as influence of extreme values

Graphs

• Bar Chart: The height of the bar represents frequnecy
• Scattergram: Suitable for correlational data, a dot or cross is show for each pair of values

Descriptive Statistics - Level of Measurment

Nominal: Comes in Categories e.g. Types of Newspapers

Ordinal: Comes in Ranks e.g. League Tables, Grades

Interval/Ratio: Fixed scores in regluar intervals e.g. Average Temp in a Month

Test of Difference: Mann-Whitney and Wilcoxon

Mann-Whitney: Difference, Independent Groups and Ordinal Data

• The calulate value of U must be equal or less than the critical value
• The results are significant at the 5% level
• So alternative hypothesis can either be rejected or accepted depending if the results are significant at the 5% level 

Wilcoxon: Difference, Repeated Measures and Ordinal Data

• The caluculate value of T must be equal or less than the critical value
• Results must be significant at the 5% level
• alternative hypothesis must either be rejected or accepted 

Parametric Tests of Difference: Unrelated and Related Tests

Unrelated T-Test: Difference, Independent Groups and Interval Data

• Normally distributed sample
• Standard Deviation in both groups will be similar
• The critical value for a two-tailed test at 0.05
• T must be equal to or more than the critical value

Related T-Test : Difference, Repeated Measures and Interval Data

• Normally distributed sample
• Standard Deviation in both groups will be similar
• Value of T must be equal to or more than the critical value

Test of Correlation: Spearman's and Pearson's

Spearman's Rho: Correlation and Ordinal/Interval Data

• The caluclated value of rhi must be equal to or more than the critical value
• Results must be significant at the 5% level, for the alternative hypothesis to be accepted

Pearson's R: Correlation and Interval Data

• The caluclated value of R must be requal to or more than the critical value
• The results must be significant at the 5% level for the alternative hypothesis to be accepted

Test of Association: Chi-Squared

• Difference or Association and Nominal Data
• Degrees of freedom: (rows-1) x (colums -1) so (1-1) x (1-1) = 1
• For significance to be shown, the calculated value of X²  must be equal to or more than the critical value 
  
• Results must be significant at the 5% level for the alternative hypothesis to be accepted

Designing Your Own Study

1. Hypothesis: State your hypothesis and then say whether it is directional or non-directional and why you have chosen this type

2. Independent and Dependent Varibales: Write out succint, Operationlised Varibales

3. Experiemental Design: State whether you will use a repeated measures, independent gorups or matched particpants design and why this is appropriate

4. Sample: Note what samping technique you would use and what sample of people you will take

5. Experiemental Method and Procedure: State what type of experiement you are using and how you will carry out the study in practical terms

6. Materials: Note any materials you need to use

7. Results: State the descriptive statistics you will use