Embed Size (px)
Citation preview
Unit G544
Research Methods in Psychology
Name: __________________
1
Task Time…Complete the table.
Study Method usedLotus and Palmer(eye witness testimony)
Baron-Cohen(autism)
Savage-Rumbaugh(kanzi)
Samuel and Bryant(conservation)
Freud(Little Hans)
Bandura et al(imitation of aggression)
Dement and Kleitman(dreaming)
Maguire(taxi drivers)
Sperry(split brains)
Haslam and ReicherBBC prison study
Milgram(obedience)
Piliaven et al(subway Samaritans)
Rosenhan(sane in insane places)
Thigpen and CleckleyMultiple personality
Griffiths(gambling)
2
Glossary of Psychological Terms
Ethics
Aims
Independent variable
Dependent Variable
Operationalise
Alternative Hypothesis
Null Hypothesis
One-tailed hypothesis
Two-tailed hypothesis
Independent measures design
Repeated measures design
Order effects
Counterbalance
3
Matched pairs design
Laboratory experiment
Field experiment
Quasi/Natural experiment
Demand characteristics
Correlation
Positive correlation
Negative correlation
Observation
Inter-Observer reliability
Test-Retest Reliability
Time sampling
Event sampling
Self report
Open questions
4
Rating Scales
Closed questions
Social desirability bias
Response set bias
Target Population
Opportunity sample
Random Sample
Self-Selected Sample
Confounding Variable
Ecological Validity
Internal Validity
Mundane Realism
Reliability
Quantitative research
Mean
5
Median
Mode
The Research Methods
There are four research methods that you must know.
The four methods you need to know about are listed below. Write yourself a definition of each including key terms which will help you differentiate them:
WARNING!! NOT ALL STUDIES ARE EXPERIMENTS!!
Research methods
Each research method has its own strengths and weaknesses. For each method identify one strength and one weakness in the following chart.
6
Experiments
Self-reports
ObservationsCorrelations
Method Strengths Weaknesses
Laboratory
experiment
Field experiment
Quasi experiment
Participant
observation
Naturalistic
observation
Structured interview
Unstructured
interview
Questionnaire
Correlation
7
Each method can be assessed in terms of validity and reliability.
Complete the following table with an evaluation of each method.
Method Reliability (e.g. how it
might be tested in this
method)
Validity (e.g. how it might be
compromised by this method
Laboratory experiment
Field experiment
Quasi experiment
Participant observation
Naturalistic observation
Structured interview
Unstructured interview
Questionnaire
Correlation
8
Questionnaires and Interviews
Why do questionnaires?
They allow us to:
1. gather large amounts of information2. in relatively short time3. in a relatively cheap way.4. Good for qualitative information about feelings/attitudes
SamplingQuestionnaires would use the same sampling techniques as experiments:Volunteer or self selected (through adverts)Opportunity (using students or those available to you for example in a shopping centre)Random (use a computer programme to select participants randomly through the voters register from the whole of the UK).
Questionnaires and dataData collected can be quantitativeData collected can be qualitativeData collected can be a combination of both qualitative and quantitative To get quantitative data add rating scales to your questions!Likert scales ( 1 – 10)Sematic scales (strongly agree, agree, disagree; never, sometimes, always etc)
Strengths of scalesScales make a qualitative measure like a feeling into a quantitative measure – this makes data between individuals easier to compare and we can use statistical analysis on the data to see if the results are significant.
Weaknesses of scalesPeople always tend to use the middle of the range ( so a 5 out of 10 or a ‘sometimes’) when giving their answer which does not enable us to see real differences between individuals.A 5 on a scale of 1 – 10 for one person represents something different to a 5 for another person. Think of how you might represent pain on a pain scale?
Qualitative questionsAdvantages are the detail and richness of the data that can be obtained.The downside is you have to analyse it – that can be difficult when people use lots of different words to describe their feelings. This makes data between individuals difficult to compare.
Open and closed questions
9
To get lots of rich data you can ask open questions:Tell me about when you last felt unhappy?To make qualitative questions easier to analyse you can ask closed questions like ‘do you feel bad’ Yes or No?
Factors which affect validity The length of the questionnaire (too long causes apathy – people may not care how accurate their answers are).The wording of each question (if any question is ambiguous, or you ask leading questions, your questionnaire will not be valid).Demand characteristics (they answer in such a way to please the researcher)Social desirability bias (they do not tell the truth for fear of embarrassment)
Improving validityHide the purpose of your questionnaire with distractor questions to avoid demand characteristics.Keep the questionnaire confidential with no names to encourage participants to be honest.Do your questionnaire on line – people are more honest to a computer! Do a pilot study to check your questions are clear and measure what you set out to measure.
Making your research more reliableMore subjectsRepeat the questionnaire at a later dateMeta analysis (compare results of different studies and see if there is a correlation of everyone's findings)Triangulation (do a lab experiment and an observation study and compare results of all three
Being ethicalDo not take names. Keep the participants confidential.Make sure they can fill out the questionnaire in private.
10
Types of self reports
Identify types of interviews from studies that you have covered. (Hint: Kolberg’s moral development)
Group work: Think of 3 different research aims and then write out three questions that you could ask
Aim
Question 1:
Question 2:
Question 3:
Aim:
Question 1:
11
Question 2:
Question 3:
Aim:
Question 1:
Question 2:
Question 3:
Types of Questions used
Identify the different types of questions that can be used and the advantages and disadvantaged
Type of question
Advantages Disadvantages
12
In Groups Compare interviews and self reports, think about the advantages and disadvantages of both
Details Advantages Disadvantages
Details of interviews
Details of self reports
Activity Time…….Measuring attitudes: scales
Qualitative or quantitative data? In their research, psychologists collect data. These data are the evidence against which psychological theories will be tested. But psychologists can do their research in many different ways and, consequently, data can take many different forms. Broadly speaking, however, all data fall into one of two categories:
• Quantitative data are numbers. They describe quantities. Quantitative data arise from questions like ‘how much?’, ‘how far?’, ‘how often?’, and ‘how long for?’
13
• Qualitative data are non-numerical. They describe qualities. Quantitative data arise from attempts to describe things and are often verbal although they can take other forms (e.g. pictures).
Many studies use both types of data to make sense of the issue being studied.
Strength Weakness
Qualitative
Quantitative
14
Quantitative measures of attitudes Psychologists who want to produce quantitative data about people’s attitudes often use attitude scales. These consist of a series of standardized questions (or scale items) the person (or respondent) answers, often by means of a pen and paper questionnaire. Each scale item gives a numerical result depending on how the respondent answers, and the results of all the scale items can be combined to produce an overall score that represents the person’s attitude towards whatever the researcher is interested in.
There are several ways of creating scale items. Two that you need two know about are:
• Semantic differential scales
• Likert scales
Why might a psychologist prefer quantitative data? Why qualitative? Which do you think would be better if we wanted to assess people’s attitudes?
15
Surveys: questionnaire design
Ask a silly question… The following questionnaire has been put together to find out about attitudes to drinking alcohol and underage drinking amongst a sample of 14 to 15 year-old school pupils. They did it in a hurry, however, and weren’t really thinking properly. Consequently, it is unlikely to elicit valid data. Have a careful look at their questionnaire design and highlight the problem areas.
Underage Drinking Questionnaire Name ________________________________________________ Address ______________________________________________Age _____ Sex _____ How often do you drink alcohol? _____________________________________________________________________ How many units do you drink a week? ______________________________________ Do you binge drink? Yes No Why do you drink alcohol? (1) because it’s fun to get drunk (2) because all my friends do it (3) because it makes me confident (4) other Do you understand the health risks of drinking and why do you still do it?
• Identify some flaws with the questionnaire and explain how they might affect the validity of the data.
• How could the questionnaire be redesigned to avoid the problems you have identified?
16
• How might a pilot study have helped to improve the design of this questionnaire?
Design your own Questionnaire or interview below. It can be on any topic which is ethical. If you can not think of a title then do “Attitudes on Speech Day at Denstone College”. Have a good look at the mark scheme first.
Sampling Methods
What do these key terms mean?
Target Population:
Sample:
Representative:
Generalisable:
Bias:
TASK: What is the target population in the following studies?
a. Research studying if people with Autism can see the world through another person's point of view.
b. A study to measure if children (age 5 years and 8 years) can tell that if plasticine is the same mass if it changes shape.
17
c. Research which measures if taxi drivers (with lots of navigational experience) have structural changes to their brain.
Sampling
From your previous knowledge match the definitions of each sampling technique. Add any
other sampling techniques you know e.g. stratified
Sampling
technique
Definition Advantages Disadvantages
Random Participants volunteer to
take part in research,
e.g. answering advert
asking for participants
Opportunit
y
Participants each have
an equal chance of
being selected, e.g.
names in a hat.
Self
selecting
Participants are used
who happen to be
available and meet
research criteria, e.g.
class of psychology
18
students
Task: For each of these situations, identify the sampling technique and comment on whether it is likely to produce a representative sample.
1. Respondents are recruited for a survey by asking passers by in the street if they would mind answering some questions.
2. Gender differences in superstition are investigated by propping a ladder against a wall and seeing who walks under it and who walks around it.
3. A sample of A – Level students is chosen by putting the names of all A - Level students into a hat and drawing out 20 of them.
5. A student recruits 15 male and 15 female students from her college canteen to take part in an experiment on memory.
6. A researcher recruits 50 undergraduates to investigate gender differences in British driving behaviour.
6. Signs are put up in games arcades in Stoke asking for volunteers for a study on gambling addiction.
7. Names are selected by random number generator from the British electoral role for a national survey on people’s current opinions about the economy.
19
Redesign three of these examples to use a different technique, and explain the choices you have made.
Complete the table below
Scenario Target
population
Sampling
technique
Rationale for
technique
Weaknesses of
technique
To investigate
children’s IQ
before and
after a course
of vitamin
tablets
To investigate
teenagers use
of the Internet
An
investigation
into
similarities 20
and
differences in
successfully
and
unsuccessfully
married
couples
An investigation into the quality of care in home for the elderly and the fees charged
Variables
What do these key terms mean?
Variables:
Independent Variables:
Dependent Variables:
21
TASK: Now identify the IV and DV in the following:
1. A psychologist wants to investigate whether students who complete their 4 hours of independent study per week do better in the psychology exam than those students who only complete 1 hour per week...
IV =
DV =
2. An experiment to see if recall on a memory test is affected by time of day
IV =
DV =
3. Does drinking coffee whilst revising improve exam results?
IV =
DV =
4. An experiment to investigate the effects of fatigue on reaction time
IV =
DV =
Operationalising
You need to know the word operationalising (to operationalise).
This refers to the process of:
Concept being
measured
How to operationalise (state precisely how you are going to measure it)
Memory – is it better in the morning or
IV - be precise about it – time of day. AM/PM – when?
22
the afternoon? ---------------------------------------------------------------
---------------------------------------------------------------
DV - Measure it get Ps to learn a list of words
Be precise(how many words) ---------------------------
Be more precise (what sort of words?) ----------------
Be even MORE precise (are you timing them) ----------
TASK: For the four studies briefly described above, say how you would operationalise the IV and DV
1.
2.
3.
4.
Extraneous and Confounding Variables
What do these key terms mean?
Extraneous variables:
Confounding variables:
23
Aims and Experimental Hypothesis
Whenever psychologists carry out a study they must start with a general aim. This is a statement of what is going to be studied. It is not a prediction of expected findings and does not need operationalising.
TASK: Read these descriptions of research and write an appropriate aim for each:
1. One group given cheese before bedtime, another group given nothing before bedtime and asked to count the number of nightmares they have during the night
2. One group given coffee before an exam, another group given caffeine free coffee before their exam. Their results are then compared
3. Men and women are observed driving and their average speed is noted
What do these key terms mean?
Alternate Hypothesis:
Null Hypothesis:
One-tailed Hypothesis:
24
Two-tailed Hypothesis:
Significance:
For your exam you need to be able to write each of these types of hypotheses correctly. Each exam question tends to be worth 4 marks so you need to get it right every time!
TASK: Think about this experiment: ‘Will students who do 4 hours independent study per week do better in the exam than those who do 1 hour per week?’
1. Identify the IV (both conditions) and the DV
IV =
DV =
2. Make sure that you have fully operationalised these (stated precisely how they will be measured).
IV =
DV =
3. Turn it into a non-directional hypothesis:
4. Turn it into a null hypothesis:
25
TASK: Now try making pairs of hypotheses for the following research questions...and then make them all the opposite direction.
1. Aggression levels in boys and girls
2. Playing driving games and the effect on actual driving skill
3. Reading and ability to complete crosswords
Sampling, design and counterbalance
Task time……
Draw a mind map about samples, remember to include advantages and disadvantages.
26
From your AS or A2 notes write out two hypothesis. Consider the above hypothesis and choose the type of sample you would use and why, then consider the effect on the generalisability of your results based on the type of sample that was used
1. Hypothesis
Type of sample:
Because
27
Effect on generalisablity:
2. Hypothesis
Type of sample:
Because
Effect on generalisablity:
Ethical issues in Psychology What are ethical issues? When we looked at research into obedience to authority, we asked whether there were any factors (such as sampling, ecological validity and demand characteristics) that affected how much we could trust it. Ethical issues are another way of assessing and evaluating psychological research studies, but they are not quite the same. Whereas issues of validity concern the quality of research in terms of whether the result can be trusted, ethical issues relate to whether a research study was morally acceptable. Ethical issues arise when there is a conflict between different sets of values that relate to psychological research. For example, on the one hand we may believe that we should be honest to people because to lie to them demeans them. On the other hand, we know that a person’s beliefs about the situation they are in affect their behaviour in that situation. Consequently we may think it necessary to deceive people about the true nature of the
28
situations in which we study them. Clearly there is a conflict between the need to be honest and the need to deceive and that is why there is an issue to be resolved. All sorts of factors can play a part in determining the ethical acceptability of a study. We will be focusing on three. These are described below.
Issue
When does it arise?
Why is it a problem?
Informed consent
When participants’ agreement to take part on a study is not obtained, or when participants are not informed about the nature of what they will be asked to do before they agree to participate.
It could be considered insulting or demeaning to people to use them in a study without their permission. If they are not fully informed about the purposes or methods of a study, then they may agree to participate then later change their minds but find it difficult to withdraw because they have agreed.
Deception
When participants are deliberately misled about the aims of research or the nature of some aspect of the study in which they are taking part.
Again, it could be considered demeaning to participants to lie to them for the purposes of research. Additionally, deception may result in stress and other types of damage to the participants (see below). The use of deception could have a bearing on whether pps have given informed consent (see above).
Protection from harm
When participants are exposed to the risk that they will be psychologically damaged either in the short or long term. This may happen in a variety of ways, for example, if the participants are stressed, or if their self image is altered or damaged.
It is considered morally unacceptable for a psychologist to damage the participants in a research study.
Why Are Ethics Important? Ethical issues are important for two main reasons. First, the aim of psychology is to increase our understanding of human behaviour in the belief that this knowledge is useful in helping people. Many psychologists would argue that this aim is incompatible with manipulating and hurting people in order to find out how they work. Second, the ethical issues raised by many studies have damaged the image of psychology in the eyes of the general public. Consequently, members of the public misunderstand the purposes of psychology and are reluctant to trust psychologists. This could lead to a situation where we run out of participants for our studies.
TASK: Read the following brief description of studies. What are the ethical principles these studies violate?
29
1. In a busy subway, a person collapses bleeding from the mouth. The person is a confederate (an actor; part of the study) and the event is staged. Bystanders are covertly observed (do not know they are being observed) to see if they help and how long they take to help. An investigation into bystander responses to emergency situations.
2. An experimenter in a pick-up truck, with a rifle visible in the back, and a sticker on the bumper saying ‘VENGEANCE’ stops at the red lights. The experimenter does not move when the lights turn green thus blocking the traffic. An investigation into the impact of aggressive stimuli on ‘horn honking’ behaviour.
3. Participants are presented with 2,000 sheets of random numbers, asked to add up 224 pairs of numbers on each sheet, and then tear the sheet into 32 pieces before going onto the next. After five hours of this useless task, some of the participants are still working through the task (adding up the numbers & then ripping up the answers). An investigation into the power of obedience.
4. Children are taken out of their nursery by a researcher. The children observe an adult being aggressive towards a bobo doll (an inflatable doll), they are then observed to see if they act aggressively and imitate the behaviour. An investigation studying if aggression can be learnt through imitation.
5. An observer hides (covert observation) in public toilets for men and records the time taken before participants begin to urinate and the time they take to urinate. A confederate of the experimenter either stands in the next urinal to the participant or one urinal ‘away’. An investigation into the effects of invasion of privacy on arousal.
6. A participant is told they are administering electric shocks to a confederate (an actor; part of the study), when the confederate stops ‘screaming’ & falls silent some participants ‘sweat, tremble, dig fingernails into their flesh and have ‘full-blown, uncontrollable seizures’. An investigation into the impact of obedience.
How many of these studies do you think are real studies?
It is important to know how to deal with ethical issues when conducting psychological research.
Working with children
30
/6
Working with people with mental health issues
Ways of dealing with potentially unethical research:
Presumptive Consent
Predictive Consent
Debrief
Retrospective Withdrawal
Task:Remind yourself of the following studies that we have looked at recently:
• Asch (1951) – majority influence
• Moscovici et al (1969) – minority influence
• Milgram (1963) – obedience to authority
31
Each of these studies raises ethical issues of the type outlined above. Write a short paragraph about each study in which you:
• Identify which ethical issues it raises
• Illustrate each ethical issue by drawing attention to the relevant aspects of the research.
Here is an example to get you started: Asch’s (1951) study raises ethical issues of informed consent. This is because the participants were told that the experiment was about visual perception rather than majority influence. Additionally, the participants were deceived about the aim of the research and were misled into believing that the confederates were actually other participants. They were put under stress during the procedure, and afterwards those who conformed may have felt foolish and angry because of the deception. This may have caused them psychological harm.
Research methods: ethical issues
1. Audience effects on helpingResearchers conducted a field experiment into the effect an audience might have on a person’s willingness to help other people in distress. Prior research had suggested that the presence of other people makes helping less likely. They conducted their experiment in an
32
isolated area of a university campus. When a passer-by entered the area, a confederate dressed in a tracksuit ran in from a different direction. In view of the passer-by, they would fall to the floor as if injured. This was done under two conditions: (1) with no other people present; and (2) with three other confederates positioned in view of the passer-by, apparently engaged in conversation. The researchers recorded whether the passer-by would help the ‘injured’ confederate and, if so, how long it took them to do so.
2. Teenagers’ attitudes towards underage sexResearchers conducted an interview survey investigating 14-15 year-old boys’ and girls’ attitudes towards underage sex. The respondents were interviewed in single-sex groups of four or five by a researcher who asked them about their sexual behaviour and their attitudes towards their peers’ sexual behaviour. An audio recording was made of the interviews and then transcribed. The transcripts were then analysed by a group of researchers to identify common themes in the things the respondents said about sexual behaviour in their age group.
3. Gender differences in physical contact between conversational partnersResearchers conducted an observation study into gender differences in physical contact between conversational partners in public places. They conducted their observation in a coffee shop. When a pair of people came in and sat down opposite each other, the researchers observed them and recorded (1) the composition of the pair (MM, MF, FF) and (2) how often they made physical contact with each other and where (hand, arm, shoulder etc.) The observation of each pair ended when one or both left the coffee shop.
4. The effect of self talk on persistenceA research conducted an experiment in which she investigated the effect of positive self-talk on persistence in a task. She designed a test of motor skills that was impossible to complete. Participants were told that most people completed the task within about two minutes. They were divided into two groups. The first group were instructed to repeat the words ‘I can do it’ to themselves whilst attempting the task. The second group repeated a different phrase with no motivational content. The researcher timed how long each participant would spend attempting the task before giving up.
For each of these research studies:• Which ethical issues are raised by the design of the study and how?
• What would the researchers need to do to deal with the ethical issues raised by their studies?
• Write a briefing script for study 4 that would be read to participants before they agreed to take part, in order to obtain informed consent.
Self Reports
33
Procedure Task: Write a procedure on how to carry out a self report on “ Life as a sixth former at Denstone College”
Review the responses of the class and evaluate the accuracy of each method
Questionnaire Procedure
Interview procedure
Levels of measurement
Nominal level data
34
This involves simply counting the number of subjects that did one thing or another, or fall into this category or that. The numbers here refer to frequency with which something occurs.
Example:
1. Number of students in the class who have blue, brown or green eyes.
2. Of 100 people, 40 recycle rubbish, while 60 do not recycle rubbish.
Think of another two examples of nominal level data.
Data are ‘measured’ roughly by putting into categories e.g. male/female, helpful/unhelpful. It is important to note that the numbers do not stand for amounts or distance on a scale.
The numbers within each category stand for quantities in that category and these are referred to as ‘frequencies’.
Example:
Imagine we survey gender differences in musical instrument choice and get the following nominal data.
Male Female
Students who play the piano 4 10
Students who play the drums 12 2
Students who play the violin 6 6
We would use a bar graph to display these data.
Students music choices
35
Ordinal level data
With this level of measurement, items may be placed in some rank order. Ordinal numbers represent positions within a group. They tell us which is first, second and third.
Example
1. In a 100 m sprint, the winner is placed first, the next runner across the line is placed
second and so on, whatever the gap between the runners.
2. A reviewer in a hi-fi magazine might rate components according to the following star
rating:
***** Excellent, buy it
**** Good, well worth considering
*** Hmmm. Aaverage
** Pretty poor show
* Avoid at all costs
Clearly a four-star rating **** is better than a two-star rating**, however we cannot assume that the four-star rating is twice as good as the two-star rating. Indeed it cannot even be assumed that the difference between the one-and two-star ratings is equal to the difference between the three-star and the four-star rating.
Ordinal numbers indicate the position of an item in a group. Data are rank ordered in some way; they are not a measure of the amount of interval numbers.
Think of another two examples of ordinal level data. (Hint: likert scale measures)
Example: Imagine we ask year 7 and year 10 students if they like their school by rating it on a scale of 1 (don’t like it) to 10 (really like it).
We get the following data:
Year 7 students
rating
Year 10 students
rating
Student A 7 3
Student B 5 4
Student C 8 3
Student D 8 4
36
Student E 6 6
Mean rating of year 7 and year 10 students
Interval level data
An interval scale uses equal intervals.
Example:
The length of time a person takes to complete a test, or the IQ scores a person gets.
This type of data gives more than just order; it also shows how much difference there is between the first and second, the second and third etc. These equal intervals could, for example, be centimetres or kilograms or seconds. An interval scale is usually defined as a scale that has an arbitrary zero.
Example:
Imagine we test how long it takes for students to complete a memory test either with or without caffeine:
Time taken With coffee Without coffee
1–10 seconds 3 1
11–20 seconds 4 2
21–30 seconds 3 3
31–40 seconds 2 6
37
For interval level data you use a histogram or frequency polygon.
Time taken for memory test with coffee
Think of three examples of interval data. Hint: Think of equipment you can buy that
measures (e.g. a clock).
38
Levels of Data (increasingly meaningful/powerful)
Nominal data: a level of measurement where data are in separate categories.
Ordinal data: a level of measurement where data are ordered in some way.
Interval data /Ratio data: a level of measurement where units of equal measurements (a scale with equal intervals) are used e.g. minutes, kilograms, number of words recalled in a memory test or percentage score in an exam. Ratio data is on a scale, but has a true zero eg weight/height, time, distance.
Looking at the pictures above, find as many examples as you can, of each type:
Nominal data Ordinal data Interval/Ratio data
39
Now identify the level of the data in each of these examples:
1. Number of children (average age 4.5 years) engaged in type of play
Non-play Solitary Associative Parallel cooperative
8 5 17 23 6
…………………………………………………………………………………………………………………………………………………
2. Podium positions, Monaco Grand Prix 2010
Mark Webber, Sebastian Vettel, Robert Kubica
……………………………………………………………………………………………………………………………………………………
3. Rosenthal and Fode (1963) Time in minutes taken by rats to correctly find their way round a maze.
……………………………………………………………………………………………………………………………………………………
4. Holmes & Rahe devised a stress scale, with arbitrary numerical values given to a series of life events such as ‘death of spouse’ or ‘getting married’
……………………………………………………………………………………………………………………………………………………
5. Bandura’s Bobo Doll study, behaviours observed: aggressive, non-aggressive
……………………………………………………………………………………………………………………………………………………
6. Schacter& Singer (1962) attempted to estimate how emotional Pps were feeling, using a scale to measure happiness, and another to measure anger.
……………………………………………………………………………………………………………………………………………………
7. Collins & Quillian (1969) measured time taken in seconds, to decide if statements were true or false.
……………………………………………………………………………………………………………………………………………………
8. Lang & Lazovik (1963) measured phobics’ scores on a 19-point scale ‘snake avoidance test’.
40
……………………………………………………………………………………………………………………………………………………
Level of Measurement Game –
Level of Measurement Questions
Question AnswersResearchers want to find the difference between male and female estimates of stopping distances.Ainsworth’s strange situation where observers had behavioural categories to observe and tick each time they saw one in a childOn a questionnaire about social behaviour there is a question asking:On a scale of 1-10 how aggressive is your child? 1 is not at all 5 average and 10 very aggressiveParticipants have to choose ‘snog’, ‘marry’ or ‘avoid’ when shown a set of photographs of possible partnersIn a company participants were asked to indicate on a scale of 1-7 how much they felt in control of their working environment.Participants are set a puzzle they can or cannot solve. Their temperature is taken as a measure of stress.One group of participants are asked to take multi vitamins for a month then their IQ is measured before and after the month is up.Researchers want to find out what type of teaching pupils prefer so ask them to put them in order of preference. They are asked to put their favourite first and so on.Researchers want to investigate the effect of leading questions on age. They want to see if those aged 5-10 were more likely to be misled than those aged 11-15.Heartbeat was measured while participants tried to complete a puzzle they can or cannot solve. Scores on a memory test before and after taking caffeineHow many days per week is your child in day care
41
How many units of alcohol per week are consumed by males and femalesA school wants to advertise its good grades at A Level and provides this by stating how many got A’s; B’s, C’s etc.Participants are asked to rate a series of photographs on their level of attractivenessFemale and male participants are asked for their actual and ideal weight.Participants are asked to complete the Social Readjustment Rating Scale (Holmes & Rahe) and calculate their score.Participants are asked to choose between an apple or a portion of popcorn. Then asked to choose again after being given the nutritional value of each.
Presentation of data
Using the Questionnaire/ interview you designed and carried out you now need to present your data. You will need a table of central tendency, data tables and a graph.
Evaluation of your research
42
Future research, suggest alternatives designs and samples: write up ideas for improvements, to your study, type of questions, design, sample etc..
TASK: Interpreting DataAs part of an ongoing project on lifestyle choices amongst Rats, research was conducted research into whether rats preferred The Red Hot Chilli Peppers to The Rasmus. The rats were placed, individually, in glass boxes. In each box was a loudspeaker and a lever. Music was played at an average sound intensity of 100dB through the loudspeaker. The rat could turn off the music by pressing the lever. The lever was also attached to a timer, which recorded the interval in seconds between the music commencing and the rat pressing the lever.
The rats were randomly assigned to one of two conditions:Condition 1: the music played was ‘In The Shadows’ by The RasmusCondition 2: the music played was ‘Give It Away’ by the RHCPs
They conducted their experiment twice, to check reliability.Experiment 1: Results
Mean duration Range Standard DeviationRasmus 17.6s 24s 5.3sRHCPs 24.3 7s 3.8s
Experiment 2: ResultsMean duration Range Standard Deviation
Rasmus 22.9s 8s 2.1sRHCPs 25.7 5s 3.1s
1. Comment on the results of the experiments, referring to all the measures given.
2. Suggest some conclusions that the researchers may have drawn from the results of theirexperiment.
43
Here are the results from the ‘Rasmus’ condition of experiment 1:
Rat Number Duration of music (s)1 182 263 164 175 26 157 128 189 1310 14
Did the researchers use the right descriptive statistics to describe their data? Explain youranswer.
Research Methods RevisionReferring to the experiment carried out above1. Suggest a suitable hypothesis for the experiment.
2. What were the IV and DV in this experiment?
3. Suggest why the rats were randomly assigned to the two conditions.
4. Identify one variable controlled by the researchers and suggest why they needed to control it.
44
5. What is meant by ‘reliability’? Were the researchers’ findings reliable?
3. Does this experiment raise any ethical issues? Explain your answer.
Extension ActivityIf you wanted to investigate whether exposure to Barry White records had an effect onreproduction in rats. Suggest how you could carry out a correlation study to investigate this.
Observations
Complete the table below
Description Strengths Weaknesses
Naturalistic
Controlled
Participant
45
Structured
Unstructured
Covert
Overt
46
Sampling behaviour
The term sampling here does not refer to gathering a group of participants for study, instead it refers to when and how to make observations. You need to know the following two methods:
Time Sampling:
Event Sampling:
Strengths Weaknesses
Time Sampling
Event Sampling
Now it is your turn, Design and carry out a simple observation. You will need to present your findings to the class using descriptive statistics, measures of central tendency and dispersion, use data tables and graphs.
Procedure
47
Write an Evaluation
Consider reliability, validity and ethics of observational data.
Write about future research
Consider alternative designs and samples.
TASK: Measuring attitudes: watch or ask?
Finding out what people think Psychologists face the problem that many of the processes they wish to investigate can not be directly measured or observed. Attitudes, for example, are hypothetical constructs. Psychologists use the idea of attitudes to explain apparent consistencies in the ways people act towards various aspects of their world. But you can’t take a person’s attitudes out and look at them to find out what they are like – attitudes don’t have an independent existence in that way. Consequently, attitudes (like many other phenomena) can only be measured indirectly. That means using a person’s behaviour to make inferences about their psychological processes.
The secret life of Year Seven Professor Fakedata and Dr Fraud are investigating peer groups in young adolescents. Specifically, they want to investigate the attitudes of year seven pupils towards their classmates. However, whilst they agree on whit they want to find out, they disagree on how to obtain the data they need. Professor Fakedata thinks that the best measure of a person’s attitudes is what they do. She plans to carry out observations of the pupils’ behaviour towards each other (1) in their classrooms and (2) in the playground. Her view is that the ways in which the pupils interact with each other will indicate their attitudes. Dr Fraud disagrees with the professor about attitude measurement. His view is that you get a more valid measurement by asking people questions about their attitudes. He is planning to interview the students but has not decided between (1) interviewing groups of students in a classroom and (2) interviewing individual students in an office. Whichever method – observation or questioning - is chosen to measure people’s attitudes, one big problem is that different researchers may ask questions or make observations in different ways.
• Of the two locations chosen by Professor Fakedata, which do you think is the most likely to yield valid data? Why do you think this?
48
• What sorts of problems might the professor encounter when gathering and interpreting her data?
• Which of the two strategies would you recommend to Dr Fraud? Why?
• What sorts of problems might Dr Fraud encounter when gathering and interpreting his data?
• Why do you think inconsistency of this sort is a problem in psychological research?
• How could you ensure consistency between researchers’ observations or questions?
Add “Doing Qualitative Observations sheet
49
Experimental Method
From your AS notes make notes on the following, explain what they are and for the hypothesis how to write a good hypothesis, include an example or two!
Research Hypothesis:
Null Hypothesis
Operationalised Variables
50
Independent Variable
Dependant Variable
One tailed Hypothesis
Two tailed hypothesis
TASK: Is it a laboratory experiment, field experiment or quasi/natural experiment? Give reasons for your answer.
A: A stooge collapsed on a train either ‘ill’ with a cane or ‘drunk’. Helping behaviour was observed.
B: The aim of this experiment was to see if licenced taxi-drivers have a larger hippocampal volume than non taxi-drivers as this is the part of the brain thought to be involved in spatial navigation. Both groups were given an MRI scan to measure their hippocampal volume.
C: Nursery children were taken to a room with lots of toys, including a 5ft inflatable bobo doll and a mallet. An adult then either played quietly with the toys or hit the bobo doll. The children then had opportunity to play with the toys alone.
51
D: The aim was to measure if Ps with autism found it harder to identify emotions from photos of eyes than a control group (general population from Cambridge).
Experimental Design-Experimental Design Types
Specific experimental research designs refer to how we allocate participants to the different conditions of the IV.
Independent measures – two groups are used, one for each condition
Matched participants – like independent measures, but the two groups of PPs are matched to be as similar as possible (e.g. age, sex, social background, level of education etc.)
Repeated measures – one group of PPs is used to do both conditionsDescription Strengths Weaknesses
Independent Measures Design
52
Repeated Measures Design
Matched Participants
Design
Some Example Experiments1. In order to assess the effects of fatigue on reaction times, a researcher gave participants a target detection test in which they pressed a button every time a dot appeared on a screen. The time between the dot appearing and the button being pressed was recorded. The participants did the test twice, once first thing in the morning, and once last thing at night.
2. In order to compare the effectiveness of two different types of therapy for depression, depressed patients were assigned to receive either cognitive therapy or behaviour therapy for a 12-week period. The researchers attempted to ensure that the patients in the two groups had a similar severity of depressed symptoms by administering a standardised test of depression to each participant, then pairing them according to the severity of their symptoms.
3. In order to assess the effect of organisation on recall, a researcher randomly assigned student volunteers to two conditions. Condition one attempted to recall a list of words that were organised into meaningful categories; condition two attempted to recall the same words, randomly grouped on the page.
4. To assess the difference in reading comprehension between 7 and 9-year-olds, a researcher recruited a group of each from a local primary school. They were given the same passage of text to read, and then asked a series of questions to assess their understanding.
53
5. To assess the effectiveness of two different ways of teaching reading, a group of 5-year-olds were recruited from a primary school. Their level of reading ability was assessed, and then they were taught using scheme 1 for 20 weeks. At the end of this period, their reading was reassessed, and a reading improvement score was calculated. They were then taught using scheme 2 for a further 20 weeks and another reading improvement score for this period was calculated. The reading improvement scores for each child were then compared.
What You Need To Do…For each of these examples, state and explain which experimental design was used by the researchers. Then answer the questions
1. Redesign experiment 1 to use repeated measures.
2. Explain why it was necessary to use a matched participants design in experiment 2
3. Redesign experiment 5 to use matched participants.
4. Suggest why repeated measures would be an inappropriate design for experiment 3
5. There are three main problems with using a repeated measures design:The PPs get to see both conditions, and this may allow them to work out the experimental aim, producing demand characteristics. The PPs may be tired or bored during condition 2. This could depress their performance. The PPs get to practice the experimental task during condition 1. This could improve their performance on condition 2.These problems can be controlled using counterbalancing.(i) For experiments 1 and 5, explain whether any of these problems were likely to affect the outcome of the experiment.
(ii) Find out what counterbalancing is, and explain for both experiments how it would have been done.
54
6. The main problem with using independent measures is that people are different to each other (participant variables) and therefore, the DV will usually be affected to some extent by the variations between people as well as the IV manipulation. This can mask the effect of the IV, especially where the expected effect is fragile or difficult to measure. In these cases, it is often a better idea to use repeated measures.
(i) For experiments 3 and 4, suggest whether participant variables were likely to have been a problem.
(ii) Explain how the researcher in experiment 3 attempted to avoid this problem.
7. Matched participants is a good compromise between independent and repeated measures, because it reduces the effect of participant variables but avoids fatigue and order effects. However, it is not used very often by psychological researchers. Suggest some reasons why this might be.
Operationalising hypotheses
In hypotheses researchers have to explicitly say how they are going to measure the behaviour they are interested in. Consider how a psychologist could operationalise (measure) the following behaviours.The first one has been done for you.
Behaviour How it could be measured
Stress blood pressure level
adrenaline levels in blood
self-rating scale from 1 (unstressed) to 10 (very stressed)
psychometric test such as self-readjustment rating scale
observation of number of times person gets angry.
Memory
55
Pain
Attachment to parents
Criminal behaviour
Sporting ability
Intelligence
56
Not only does the dependent variable have to be defined or operationalised, but so does the independent variable. If a researcher wanted to study the effect of age on memory, how age is defined may vary. It may be defined as people over 50 and people under 50, or people over 35 and people under 35. It is up to the researcher to define how they are categorising their conditions.
In the table below write ways in which psychologists might categorise these independent variables.
Independent variable Categories
Organisation in memory Clustering words vs unclustered words
Lists of word vs scattered of words
Headed lists vs unheaded lists.
Intelligence
Social class
Sporting ability
Gender
Sociability
Personality
57
Research Design
For each of the following research situations, design an experiment to investigate it. In your experimental plan…You must…_ Give a brief outline of the experiment_ State what are the IV and the DV_ Suggest a suitable alternative hypothesis for the experiment_ Explain how your IV and DV would be operationalised_ Explain whether (and why) your hypothesis is directional or not_ Suggest some suitable controls_ Explain which experimental design you would use_ Explain the importance of the controls you have selected_ Outline the advantages and disadvantages of the experimental design you have chosen
Example 1An investigation to find out whether listening to music affects the quality of college work.
Example 2An investigation to find out whether one or other sex is more superstitious
Example 3An investigation to find out whether studying psychology makes you happy
Example 4Another investigation based on your own ideas…
Control of extraneous variables
What is meant by the term extraneous variables?
Think of a study. Who conducted the study?
What is it about?
58
Can you identify cause and effect? Explain
Could you replicate it?
If not why not?
Does it have ecological validity?
Demand characteristics?
Validity and Reliability
These terms are incredibly important in Psychology. You WILL be using them in the exam, BUT you cannot just throw the terms in and expect to pick up marks. You MUST use them appropriately and show the examiner that you understand what they mean.
What do these key terms mean?
Internal Validity:
Ecological Validity:
59
Internal Reliability:
External Reliability:
Factors affecting the validity and reliability of research
What do these key terms mean?
Experimenter Bias:
Demand Characteristics:
60
High
Low
Reliability
Validity
Validity
Reliability
Social Desirability:
Checking the validity of research
Face Validity:
Concurrent Validity:
Pilot Study:
Checking the reliability of research
Split-half method:
Test-retest:
Inter-rater reliability:
Lack of experimental control of extraneous variables
Single Blind:
Double Blind:
Artificiality (low mundane realism)
61
Ways of dealing with this:
62
TASK: Look at the scenarios below and complete the answers in the columns
Scenario Evaluate the validity of the DV in
this study
Evaluate the reliability of the DV
in this study
Suggest an alternative way of operationalising
the DV
A psychologist wants to find out if there is a difference in the memory ability of males and females in an experiment testing their ability to recall words from a list of 20.
A psychologist has arranged to test concentration of children in an after-school club to see if they complete a jigsaw faster when they are in a noisy room (their usual play area) than a quiet room (a room where they are alone with a supervisor).
Researchers are interested in finding out what makes people ‘attractive’. They are asking Ps to rate photos of 20 males and 20 females for attractiveness on a scale of 1-10 where 1 is very unattractive and 10 is very
63
attractive
Add “Doing Experiments” flowchart
64
Materials and procedures
Devise a research questions
Write out a proposal for your procedure
65
Identify the IV:
DV:
How replicable is the procedure?why?
How are variables measured?
Level of data to be collected and strengths and weaknesses of this type of data
Ethics?
Outline the main ways to present data
66
Outline the data gained from your own experiment
Conclusions drawn from your experiment
Evaluate your experiment (Discuss strengths and weaknesses of method used, Think: Cause and effect, Extraneous variables and controls, Validity and reliability of measures & Ethics)
67
Identify any extraneous variables
Any controls
Consider alternative design and samples and suggest why they would be appropriate.
68
Task: Data InterpretationFakedata & Fraud (2001) wanted to find out whether Psychology students were better looking that non-Psychology students. 20 Psychology students and 20 control participants were independently rated for attractiveness by a female psychology lecturer on a scale of 1 (least attractive) to 5 (most attractive). The ratings are given below.
PP number Psychologists Non psychologists1 2 22 5 23 4 44 3 35 3 26 4 57 2 38 2 49 1 410 4 311 3 212 3 213 2 114 4 215 3 416 1 217 5 318 4 319 5 520 5 2
1. State a suitable null hypothesis for this investigation.
2. Calculate the mean attractiveness ratings for the two groups in the study.
3. Draw a frequency histogram to illustrate the results of this study. Make sure you give it atitle and label the axes.
69
4. Explain, by referring to the frequency histogram, why the mode is an unsuitable measureof central tendency for this data set.
5. Suggest whether the researchers would be able to accept their alternative hypothesis.
6. Identify and explain one design flaw in this investigation and suggest how it could becorrected.
70
In psychology we use statistics to tell us if our research results were simply a fluke or if our IV really had an effect on our DV
Or
If we did a correlation, if there really is a significant relationship between our variables (i.e. tidiness and parental strictness) or is it just a fluke
Imagine you have just done an experiment to see if drinking caffeine affects time taken to solve a crossword puzzle.
Your IV is caffeine or no caffeine Your DV is time taken to solve the puzzle
You work out the mean scores (average) of each group and put the data on a bar chart
Bar Chart showing the Mean scores of time taken to solve the crossword puzzle
71
‘eyeballing’ the data, it looks like there’s a difference but we want to be sure that our results are not just a fluke...
‘eyeballing’ the data, it looks like there’s a difference but we want to be sure that our results are not just a fluke...
In fact we want to be 95% sure our results were not due to chance. So we do a statistical test to find out.
In psychology we have decided that being 95% sure our results are not due to chance is the percentage that we will use. There are a number of reasons for this...
For example we could set our percentage lower and say that we only wanted to be 80% sure our results were not due to chance. However, if we set our percentage lower we could end up saying our results are significant when they really aren’t because there’s such a larger margin for error – 20% possibility that our results were due to chance (this is called a Type One error)
Similarly if we wanted to be VERY sure our results were not due to chance we could say we wanted to be 99% sure our results were not due to chance (only a 1% or less margin for error). However, we then have a possibility that we say our results are not significant when in fact they are because we’ve set our level of significance too high – been too stringent (this is called a Type Two error).
Now we have to actually do a statistical test.
If we are doing a correlation we would use Spearman’s Rho If we are doing an independent measures design experiment (like my caffeine
experiment0 we would use a Mann Whitney U test
Imagine you have collected your correlational data (for example on parental strictness and tidiness) and completed all your calculations on your Spearman’s Rho equation (or got the statistics package on the computer to do it for you - see iLearn/NLN materials). You will end up with a number.
For example rs = 0.654
You now need to compare this number against a table of numbers (values) which will tell you if your results are statistically significant or not...
72
So, we stick with being 95% sure our results are not due to chance. Put another way this means that there’s a 5% possibility or less that our results are due to chance.
You can write this a number of ways and you need to know this for the exam!!
Most commonly it is shown as a decimal p≤0.05 this means there is a 5% chance or less that the results were a fluke.
In other words we are 95% sure our results are significant and NOT due to chance.
So, we stick with being 95% sure our results are not due to chance. Put another way this means that there’s a 5% possibility or less that our results are due to chance.
You can write this a number of ways and you need to know this for the exam!!
Most commonly it is shown as a decimal p≤0.05 this means there is a 5% chance or less that the results were a fluke.
In other words we are 95% sure our results are significant and NOT due to chance.
This is called the observed value or calculated value because you have just calculated it!
This is called the observed value or calculated value because you have just calculated it!
In the exam you might find a table that looks something like this (you will be given the observed value)
Table of critical values for Spearman’s test
p≤0.05 p≤0.025N=20 0.380 0.447
(
The observed/calculated value of Rho must be equal to or greater than the critical/table value to be significant)
You should also be told how to compare the observed and critical values in the exam.
So we need our observed value of rs = 0.654 to equal or exceed the critical value of 0.380.
As 0.654 is greater than 0.380 we can say our data is statistically significant !!!
So we will ACCEPT our alternate hypothesis and REJECT the null hypothesis. We can say there is a 5% possibility or less that our results are due to chance
Finally...if results are NOT statistically significant i.e. your observed value is less than the critical value (for a correlation), we would write this as p≥0.05 meaning.......................................................................................................
73
Remember our level of significance is p≤0.05 so this is the number you need to choose. This number is called the
critical value
Remember our level of significance is p≤0.05 so this is the number you need to choose. This number is called the
critical value
5% chance of it being a fluke5% chance of it being a fluke
N= No of ppsN= No of pps
≤
2.5% chance of it being a fluke2.5% chance of it being a fluke
Flowchart for choosing a test
4. Complete the flowchart to show which statistical test should be used to analyse data.
74
Choosing the Correct Statistical TestThere is a range of different statistical tests available. You need to apply the correct one toyour data, or the result will be meaningless. In order to choose the correct test, you need to correctly answer three questions:_ What are you testing for?_ What type of data do you have?_ Do you have related or unrelated data?
Decision 1 – what are you testing for?This is an easy decision to make, because it’s already stated in your hypothesis. You are either testing for a difference between two sets of data, or a correlation between two variables.
Decision 2 – what type of data do you have?There are three types of data you could have.
Nominal You have been counting things and putting them into categories (e.g.male/female). Your data are frequencies; they cannot be arranged into rank order, and there is no scale of measurement
Ordinal You have been measuring things and have data that can be put in rank order of size. However, there is noscale of measurement with equal units.
Interval You have been measuring things on a scale that has equal units (e.g. seconds, centimetres, degrees Celsius).
What type of data do you have?
Decision 3 – are the data related or unrelated?
Related Data Your data consist of paired observations. That is, each datum in set one has a corresponding datum in set two. In practice, this usually means that each pair of measurements came from the same person (either a correlational or repeated measures design).
Unrelated Data Each datum is a unique observation (an independent measures design was used). Are your data related or unrelated?
75
“Statistical Tests on Data” Add handout
76
77
78
79
80
Correlations
Whereas experiments look for a difference between two conditions, a correlation is a way of measuring the relationship between two variables to see if a trend or pattern exists between them.
WARNING!! It is very important that you do not use experimental terminology when talking about correlations. Correlations assess the relationship between variables (variable 1 and variable 2) but there are no independent variables involved as nothing is manipulated.
What do these key terms mean?Positive Correlation:
Negative Correlation:
TASK: Look at the examples below and decide whether the correlations are positive or negative:
1. The more students revise, the higher the exam grade they achieve
2. The more lessons that students miss, the lower their commitment to learning grade
3. The more apples people eat, the less fillings they have
4. The lower your expectations, the worse you do in exams
5. The less water you drink the more likely you are to have a headache
81
Writing correlational hypothesis
Remind yourself of these terms and link them to correlations:
Directional:
Non-directional:
Null:
In essence, writing correlational hypotheses is the same as we practised for the experimental method except you change the word ‘difference’ for the word ‘correlation’.
TASK: Write correlational hypotheses for these scenarios:
1. The relationship between aggressive behaviour and watching aggressive TV
2. The relationship between car speed and accidents
3. The relationship between rainfall and hosepipe bans
4. The relationship between hours of sunshine and time spent on the beach
82
Presentation of Correlational Data
The only visual display that is suitable for a correlation is a scattergraph. Also the only time you can use a scattergraph is to represent correlational data – do not use it for anything else. Dots are plotted on the scattergraph to represent each pair of scores.
Remember that you are plotting the two variables against each other, you are NOT plotting participant numbers!!
Draw a sketch of a graph illustrating a positive correlation
Draw a sketch of a graph illustrating a negative correlation
Draw a sketch of a graph illustrating no correlation
83
Strengths of correlations Weaknesses of correlations
84
‘There will be a significant positive correlation between ambient temperature and ice cream sales’ [Directional]
‘There will be a significant correlation between the amount of coffee people consume and their performance on a reaction time test’ [Non-directional]
‘There will be no significant correlation between hair colour and IQ score’ [Null]
TASK: Fill in the blanks to compare experiments and correlations
Advantages Disadvantages
Experiments________ can be determined because we observe the effect of an IV on a DV
The procedures can be ________ again relatively easily, which means that the findings can be confirmed.
Artificial; a contrived situation where participants may not have behaved ____________.
May lack ____________because cannot generalise from the experimental setting to other (real-world) settings.
Studies using correlational analysis Can be used when it is not
possible to____________ variables.
If no correlation can be found, a __________ relationship can be ruled out.
People often misinterpret correlations and assume that ____________ has been found, whereas this is not the case.
There may be other unknown ____________that can explain why the co-variables being studied are linked.
Why would we use correlations rather than experiments sometimes?
List your reasons...
85
Task: Design a correlation:
86
Evaluate correlational design (Think validity, reliability and ethics)
Alternative designs and samples. Discuss what effect these would have on the results.
87
Interpreting correlations
How can we explain these relationships?In the table below are some real correlational relationships that have been established through research. For each one, suggest the underlying nature of the relationship. This first one is easy, but the remainder are a bit more challenging.
Correlation Underlying nature of relationshipA moderate positive correlation between thenumber of times a student is praised by ateacher and her/his final mark in that subject.
A weak positive correlation between thenumber of violent films a person has watchedand their own level of aggression.
A strong positive correlation between icecream sales and accidental deaths bydrowning (in the UK).
An almost perfect correlation between thenumber of Christian churches in a city and the number of crimes committed each year.
88
Doing Correlations
Decide on an aim
What is the general question you wish to answer? Which factor do you think affects which other factor, and how?
Formulate a hypothesis
What specific prediction can you make about your results?
Operationalise the variables (need to be things you can physically measure)
Measure variables
How will you measure the two variables?
Sample
Who is your sample?
Type of sample and details
Collect data
Run the experiment: exactly what will you do?
Remember Ethics
Remember extraneous variables
Remember controls
Analyse data
Draw a table of results
A scattergraph (Hint: remember labels)
Use statistics to make sense of the data you have obtained
Accept or reject hypothesis
Spearmans Rank
89
Accept or reject hypothesis
Decide whether your data are in line with the prediction you made about the results.
Planning the 13+6
Formulate a hypothesis
What specific prediction can you make about your results?
Operationalise the both variables. (these need to be things you can physically measure)
Measure variables
Experiment Correlation Observation Self report
Identify the IV & DV
How will you measure the two
variables?
Identify the two variables.
How will you measure the two variables?
What are you going to measure?
If it is a controlled observation identify the
IV & DV
Sample
Name sampling method (Opportunity, random, self selecting) Explain how they were selected.
Who is your sample? (e.g. 20 college students aged 18 years)
Allocation of sample to groups.
Procedure
Exactly what will you do? THINK: What, Where, When & How
Experiment Correlation Observation Self report
Design: Independent/Repeated
measures? Matched pairs? Counter balance?
Detail of conditions
Standardised instructions
timings
Materials needed
Controls
DO NOT DESIGN AN EXPT
Details of what to do?
Standardised instructions
Materials needed
Controls
Event or time sampling
Coding scheme & Criteria
Materials needed
Conditions
Timings
Details of how administered
Standardised instructions
Materials needed including examples of questions
Split half method?
90
Identify and control extraneous variables
REMEMBER : What, Where, When & How
Data Collected
What type of data is collected? Ordinal/Nominal/Interval
Outline method of scoring data clearly (1 mark per correct answer)
Ethics
Not asked personal data, right to withdraw, Informed consent, confidentiality, over 16 years or parental concent, debrief participants, avoid distress, harm or embarrassment.
IS YOUR DESIGN PRAGMATIC??
91
Procedure for an experiment.(Suggested writing frame) “ASMILE”
STRUCTURE USEFUL
STARTERS
USEFUL
VOCABULARY
CONTENT
AIM With the aim of investigating…
Clearly state the aim of your study (wide and not a prediction)
SAMPLE The researcher used… Participants
Target population
Opportunity
Number of Ps
Demographics (age, gender etc – at least 2 of these)
Target population
METHOD A ______ experiment with a(n) _________ design was used
Laboratory
Field
Natural
State the type of experiment and the design used.
How Ps were assigned to conditions
INSTRUCTIONS Ps were given standardised instructions…
Distracter questions
Recall
Recognition
Full, step by step details of what Ps were told & details of the task they carried out including any materials used, how the data was recorded
and any timings used.
LOCATION The research took place…
Where Ps were, if they were alone or in groups, time of day
ETHICS/ EVALUATION
Ps were debriefed…
A strength of this procedure is…
Right to withdraw
Control, Causal
Valid, Reliable
Ecological validity
Order effects
Individual diffs
Explain how any ethical issues were dealt with – how they were
debriefed (thanks, aim) & offered the right to withdraw data
Your evaluation must be balanced –
strengths and weaknesses! Use ‘signposts’ (e.g. ‘however’ or ‘on
the other hand’) to show the examiner what you are doing. Use
terminology but make sure that you demonstrate your
understanding of what the terms mean rather than just throwing
92
A weakness of this procedure is…
Atypical
Generalisable
Representative
Demand characteristics
them in.
93
