research methods completed Flashcards

Question

what is social desirability bias

Answer 1

where participants feel that they will be judged in some way the reseacher to present themselves in more desirable light

Answer 2

1. single blind technique (participants unaware of condition they are in as conditions are made to look identical) 2. deception- lead them on to believe its about something else (aim) 3. change type of experiment to field to field of natural ( so ppts don't know they are being studied)

Answer 3

when investigator influences the outcome of the study subconsciously convey his want for certain findings to the participants more enthusiastic with wanted group, not giving clear instructions, give more detail to one condition than other

Answer 4

standardised instructions- read to both participants ie script double blind technique- ppts and researcher don't know what condition is what, made to look similar often used using third party researcher to conduct research

Answer 5

1. situational variables 2. investigator effects 3. participant reactivity 4.participant variables

Answer 6

way of overcoming participant variables in independent measures design 1. assign each participant a number 2. write number on slip of paper, fold and put into a hat 3. first number drawn goes to condition a, then the next goes to b, then continues until all participants are signed a condition

Answer 7

must be written as a script 1. step by step procedure for everything from beginning to end 2. tasks which participants are asked to do are clear 3. timings if possible 4. ask for any further questions before start

Answer 8

consistency in results and if it can be replicated or not, if a different researcher used it

Answer 9

control, ie using lab setting, standardised procedures and operationalisation replicability is only established by replication of the results by another researcher.

Answer 10

accuracy of the findings/ measures of the study and whether the researcher is measuring what they intended to

Answer 11

accuracy within research itself and the way it has been conducted (are the results due to manipulation of iv or have we measured the effects of extraneous variables or confounding variable) lack of control if there are ev's lowered internal validity

Answer 12

control ev's

Answer 13

whether the results can be applied outside of research situations 'generalisability' extent which research can be applied to other circumstances will the findings be an accurate representation of what behaviour is like in other situations temporal, ecological, population

Answer 14

1. temporal validity 2. ecological validity 3. population validity

Answer 15

can the findings be generalised to other periods of time

Answer 16

can the findings and conclusions of the study be generalised to other settings

Answer 17

can the findings be generalised to other groups of people

Answer 18

participants experience both or more conditions of the experiment, takes the form of participants being tested before and after the manipulation of the iv

Answer 19

controls for individual differences between participants as they take part in both conditions, reduces ev's

Answer 20

demand characteristics, both conditions being taken part of reduces internal validity order effects, participants performance affected by performance in first to practice, boredom or fatigue. reduce internal validity

Answer 21

involves using different individuals in each condition of the study possibly randomly assigned

Answer 22

less chance of demand characteristics compared to repeated measures, only aware of one condition, less likely to guess the aim no order effects to confound the results which you would get in repeated measures design, participants becoming skilled, fatigued, only take part in one condition

Answer 23

individual differences (participant variables) between two groups before the study begins, wouldn't get in repeated measures.

Answer 24

using different participants in each condition but firstly matching them up on a variable to control the study participants pre tested on variable, matches into pairs on similar performance. split into different conditions

Answer 25

eliminates the order and practice effects that a repeated measures design suffers from, ensure participants do not become fatigued as they don't take part in both conditions. reduces individual differences between participants as they are matched with someone who is of similar socioeconomic background for example.

Answer 26

impractical, time consuming and difficult to match up participants on specific variables. impossible to match on every variable that could possibly effect the results

Answer 27

to balance out order effects so they don't effect the performance in one condition reduces the IMPACT of order effects used for repeated measures design

Answer 28

1. split participants into two groups 2. one group does condition A and the other does b 3. the second group does condition b then a

Answer 29

lab field quasi natural

Answer 30

researcher controls as many variables as possible, control over who what where when why direct manipulation of the independent variable by the researcher highly controlled

Answer 31

high levels of control over ev's, means that we can be confident in effect on dv, increase internal validity high control can be replicated - reliability

Answer 32

artificial laboratory experiment- lack ecological validity, may not apply to real life behavior demand characteristics more likely to occur, participant more likely to guess aim

Answer 33

natural setting like local street, library iv is manipulated and behaviour observed by researcher

Answer 34

greater ecological valitiy, more relistic reduce demand characteristics, unaware being studied

Answer 35

more difficult to control variables researcher less confident iv has had effect on dv ethical issues like invasion of privacy, deception, stress and embarrassment

Answer 36

real life setting, iv occurs naturally and not directed by experimenter mostly used for experiments where it would be unethical to manipulate the iv

Answer 37

permits study aspects which couldnt be manipulated or studiesd in any other way higher ecological validity, natural behaviour in natural environment

Answer 38

litrle control over variables including iv cant be certain on cause and effect ppts cant be randomlly allocated to conditions, could be differences in ppts which could affect dv cant be replicated

Answer 39

iv occurs naturally and based on existing difference in people study of gender where males and females are compared, or age often conducted in controlled setting but not always random allocation isnt possible

Answer 40

high level of control over evs can be replicated due to controlled settings, reliability can be established

Answer 41

ppts cant be randomly allocated to conditions, could be differences in conditions which could effect dv (confoudning variable)

Answer 42

natural - independednt variable varies naturally to due to external event, researcher studies the effect this has had on the event occurred in quasi the researcher can't freely manipulate the IV as it is based on existing differences in people which cant be changed (age, gender, personality)

Answer 43

1. correlations 2. observations 3. questionnaires 4. interviews 5. case studies

Answer 44

to study relationship in 2 variables has co variables analysing data quantitative 3 types positive negative no correlation

Answer 45

numerical measure of strength of relationship between 2 variables from -1 perfect negative-----0 no correlation -------+1 perfect positive

Answer 46

a relationship between two variables in which both rise and fall together

Answer 47

a relationship between two variables in which the value of one variable increases as the value of the other decreases.

Answer 48

curvilinear correlations draw scatter graph because data gives correlation coefficient of 0 and which suggests no relationship when there is one There is a non-linear correlation when there is a relationship between variables but the relationship is not linear (straight)

Answer 49

' there will be a significant positive correlation between the ..... (covariables)

Answer 50

'there will be a significant relationship between the .... covariables'

Answer 51

'there will be no significant relationship between .... covariables'

Answer 52

can be used to study variables that can't be manipulated for practical or ethical reasons useful for further research

Answer 53

correlation doesnt mean causation - not possible to establish cause and effect could be intervening variables correlation coefficient only useful for linear relationships

Answer 54

naturalistic vs controlled covert vs overt participant vs non participant

Answer 55

occur in natural setting where behaviour would typically occur, not manipulated or controlled by researcher researcher tries to remain in background and those being observed are usually but not aways unaware of presence

Answer 56

reducing the naturalness of behaviour being studied ppts more likely to know they are being studied, ppts are less likely to know they are being studied and could be in lab

Answer 57

strength - high ecological validity and findings can be generalised to everyday life limitation- lack of control over the situation makes evs more encouraged, reduce internal calidity of the observation, replication is also harder

Answer 58

strength- high levels of control means the likelihood of evs are reduced, increasing internal validity of controlled observations replication of observer is also easier limitation- as behaviour is studied in highly controlled environment, controlled observations are artificial meaning results lack ecological validity

Answer 59

ppts unaware they are being observed, in secret

Answer 60

ppts aware being observed, often given informed consent before hand to be studied

Answer 61

dint know they will be observed removes participant reactivity increasing validity

Answer 62

ppts can give informed consent before observation and withdraw at any point if they are no longer comfortable with observation, making it more ethical

Answer 63

researcher part of group they are studying, to provide first hand account

Answer 64

strength - researcher can experience situation as the participants do; giving them increased insight which may increase validity of findings limitation- researcher may identify strongly with those who are studying and lose obejctivity some researchers describe this as 'going native'

Answer 65

when researcher stays separate from those who are being studied, only option sometimes when impractical for researcher to be involved

Answer 66

strength - allow researcher to maintain objective distance from ppts so less danger of going native limitation- lose valuable insight to be gained participant observations as they are tp far removed from people being studied

Answer 67

strength; - used in situatins when manipulation of researcher would be impractical or unethical, ie working with children or animals limitation-= - observer bias researcher pay attention to specific aspects of situation to support their hypothesis, subjective - cant establish cause and effect, no manipulation of iv

Answer 68

unstructured structured

Answer 69

researcher writes down everything that they see

Answer 70

researchers create structured list of behaviours to observe and telly each time they occur

Answer 71

strength qualitative data produced, non numerical, be rich in detail and provide insight limitation- harder to analyse, more susceptible to observer bias as research may only focus on behaviours and catch their eye as they dont have list to stick to

Answer 72

strength- less chance of observer bias as they have list of behaviours which they cant deviate from limitation- structured observations produce quantitative numerical data, whilst easier to analyse, lacks the level of detail seen in unstructured observations

Answer 73

objective, no judgements cover all possible behaviours mutually exclusive behaviours, shouldnt have to mark two categories

Answer 74

number of times that behaviour (event) occurs ie woman playing with hair

Answer 75

strength better to use if behaviour is infrequent as could be missed in time sampling limitation complex, important behaviours may be missed if behaviours occur over long period of time

Answer 76

record what ppt does in given time frame, what they do every 60 seconds for example

Answer 77

strength: lasts for long time, useful if behaviours last for long time ie- flirting so leaning in, tally it in every interval limitation of time sampling important behaviours could be missed which arent in time frame

Answer 78

observer bias can be present, may lack reliability, so we use two observers to check recording is objective to have inter observer reliability behavioural categories must be used and need to be trained in them, pilot study to check everyone uses them, observe behaviour at the same time

Answer 79

using a correlation should be positive, correlation coefficient of atleast 0.8 which suggests it is objective

Answer 80

questionnaire interviews - structured or unstructured

Answer 81

ppts reporting details of own thoughts/ feelings and behaviour to the researcher involve open or close questionnaires

Answer 82

open- produce qualitative data closed-restrict answer, yes or no

Answer 83

strength: lower social desirability bias and investegator effects, no face to face contact, researcher may unconsciously convey hypothesis through tone of voice or face expressions, researcher gone so truth and honest answers limitation response bias- responding in particular way to the questions, favour scale as they are reading questions too quick or not reading it carefully, 'yea saying' tendency to agree regardless of questions

Answer 84

likert scales rating scales fixed choice

Answer 85

statement to agree or disagree from 1-5 strongly agree to strongly disagree

Answer 86

1-5 how entertaining for example, strength of feeling of topic not entertaining to entertaining

Answer 87

tick what applies list of possible options

Answer 88

structured or unstructured

Answer 89

list of questions which requires to choose selection of possible answers, doesnt deviate from questions

Answer 90

one of number of topics to cover but questions arent pre set, less formal, flexible, in depth interviews, open questions to guide interviewee to talk in own terms

Answer 91

easily replicated as its standardised, less chance of investigator effects, consistency between interviewers and ensures reliability researcher cant deviate or elaborate on topic, frustrating leads to loss of detail and important info could be missed

Answer 92

leads to qualitative data, can expand on answers in own way, leads to greater detail and new insights into topic, researcher able to delve into answers that are interesting for research investigator effects- face to face contact could lead to unconsciously conveying hypotheses by tone of voice or facial expressions

Answer 93

interview schedule, structured, standardised minimise investigator effects be friendly and pleasant but business like not be emotionally involved, no communication of feelings to answers

Answer 94

argue key to successful unstructured interviews is trust, interviewee feel sympathetic and that there opinions are interested in by interviewer ask to clarify and develop their answers give the interviewee opportunity to develop answer and express themself freely

Answer 95

in depth detailed investigations about individual or small group of people, analyse unusual individuals or events (ie person with rare disorder) - use range of sources to build picture -may involve gathering info from friends and family - conducted over long period of time- longitudinal study

Answer 96

strength- large level of detail, qualitative and quantitative used, depth better high levels of validity, accurate picture of what is being studied limitation- main issue with case study method, only focuses on one individual or smaller group of people, cant generalise findings from case studies to whole population

Answer 97

The participants themselves choose if they want to be part of the experiment (they are ‘self-selected’)

Answer 98

strength It is a relatively straightforward way of gaining a large sample of people as participants put themselves forward. This also ensures they will be motivated to take part and less likely to drop out. limitation Because it is self-selected, it is biased, leading to an unrepresentative sample. People with certain personality types volunteer to take part in experiment- for example, Rosenthal (1965) suggested that volunteers tend to be younger, more outgoing and more confident than non-volunteers. Therefore, this sampling method can lack population validity.

Answer 99

people who are available to take part at the time and place of the study, researcher approaches to see if they agree to take part in the study

Answer 100

strength Because it is using whoever is available at the time, it doesn’t require the level of planning and preparation that many other sampling methods require. This will save the researchers time and also money! limitation Researcher bias is an issue in opportunity sampling because the researcher has control over who they choose, meaning that they may approach a certain type of person more than others, if they feel the individual may be particularly helpful or support the hypothesis the researcher is trying to test.

Answer 101

selects participants by chance from target population, every person equal chance of being selected - every name assigned to number in target population - number written on paper into hat or computer generator -pick out desired sample of number until have correct sample size

Answer 102

strength It can be more representative compared to volunteer and opportunity sampling. This means that because everyone in the target population is included in the sampling process, random samples are often more representative of everyone, leading to higher population validity limitation Needs a complete, up to date list of the target population, which in some cases is not possible so the sampling method cannot be used. E.g. it would be very difficult to draw up a comprehensive list of drug addicts in the UK. It is also more time consuming than other methods.

Answer 103

Involves taking every nth person from a list (‘sampling interval’) to create a sample. Target population ÷ desired sample size = sampling interval

Answer 104

strength The results are representative of the population, unless certain characteristics of the population are repeated for every nth person, which is unlikely. A sample gained through this method is likely to have high population validity. limitation Can still be unrepresentative. Even though the sample is chosen using the sampling interval, you could still end up with a sample that is biased e.g. more females drawn, or most of them come from a similar area or background.

Answer 105

This is a small-scale reproduction of the target population. It involves dividing a population into characteristics important for the research – for example, by age, social class, etc. These are called strata. The stratified sample attempts to reflect the proportions (of a given stratum) seen in the target population, within your sample. Importantly, the participants are randomly selected from each strata using the same strategy with random sampling

Answer 106

strength Highly representative - The aim of stratified sampling is to reproduce the traits of the target population within your sample, in the same proportions, which makes this sampling technique very representative. limitation Time consuming. The dividing of a population into strata and then randomly selecting from each one can be quite time consuming compared to the other sampling methods, suc

Answer 107

deception lack of informed consent lack of protection from harm confidentiality lack of right to withdraw

Answer 108

This should be avoided whenever possible. Participants should not be misled or information withheld from them. The use of deception prevents informed consent so the participants may be involved in research that goes against their wishes. However, there are occasions when deception can be justified if it doesn’t cause the participant undue distress, such as cases where revealing the aim would cause unnatural behaviour from participants.

Answer 109

Informed consent means that the research participants should be fully informed about the nature and purpose of the research and their role in it then allowed to agree or refuse (consent) to participate in the research. There are many situations were giving informed consent isn’t possible e.g. when deception is being used or when participants are not aware that they are part of research (e.g during a covert observation). This is also an issue because if participants are not fully informed about their role in the research they may be distrustful of psychology research in the future. Additionally, children under the age of 16 cannot give informed consent - this must come from a parent or guardian.

Answer 110

Protection of participants from harm means that research participants should be protected from undue risk during an investigation. The risk of harm should be no greater than that in ordinary life. The researchers should aim to ensure that the participants leave the experiment in the same state (both mentally and physically) as they arrived in. This is an issue as lack of such protection denies important rights to the participants. This might include embarrassment, loss of dignity or threats to a person’s self-esteem as a result of participation.

Answer 111

Participants have the right for their data to remain confidential and not be traceable back to them. Their names and other personal details should not be used within the publication of any research and they should have the right to remain anonymous throughout the whole process. Confidentiality refers to our right, enshrined in law under the Data Protection Act, to have personal data protected.

Answer 112

The subjects must be made aware, prior to the investigation, that it is their right to withdraw from the investigation at any time they choose. They should also be offered the opportunity to withdraw their results AFTER the study once they have been informed of the true nature of the research . This is called ‘retrospective right to withdraw’. Withdrawing their results means that they take no further part in the research.

Answer 113

An ethics committee has the responsibility to weigh up the costs versus the benefits of research proposals to decide whether a study should go ahead or not. Benefits may include the value of ground breaking nature of the research. Costs may be the damaging effect on individual participants or to the reputation of psychology as a whole.

Answer 114

Dealing with informed consent: Participants should be issued with a consent letter or form detailing all relevant information that might affect their decision to participate. Assuming the participant agrees, this is then signed. For investigations involving children under 16, a signature of parental consent is required. There are other ways to obtain consent if it is not possible to do before the investigation: · Retrospective consent – ask for informed consent after they have taken part in the study. The participants need to be fully informed in what they have just taken part in. · Prior general consent – gain general consent to be involved in investigations that may use deception. · Presumptive consent- rather than getting consent from the participants themselves, a similar group of people are asked if this study is acceptable. If this group agree, then consent of eth original participants is “presumed”.

Answer 115

debriefing At the end of a study, participants should be given a full debrief. Within this, participants should be made aware of the true aims of the investigation and any details they were not supplied with during the study, such as the existence of other conditions in the experiment. In the debrief, participants should also be told what their data will be used for and must be given the right to withdraw data if they wish. This is particularly important if retrospective consent is a feature of the study. Participants may have natural concerns related to their performance within the study, so should be reassured that their behaviour was typical or normal. In extreme cases, if participants have been subject to stress or embarrassment, they may require counselling, which the researcher should provide.

Answer 116

All personal details must be protected and anonymity should be maintained if details are to be kept on record. . However, most researchers use numbers or initials when writing up the investigation. Such as with the case study of HM in investigating memory. They are also to be reminded throughout the research that their data will be protected.

Answer 117

· What participants will have to do in the research, which should be detailed enough so they can give ‘informed’ consent e.g. tasks required, time required of them, any anticipated distress. · Give the participants the right to withdraw. · Ensure confidentiality. · Ask the participants whether they consent to taking part · Ask whether they have any questions about the study.

Answer 118

· The real nature of the study with aims. · Gaining consent now that participants know the real aim · Allowing participants to withdraw their data if they are unhappy with having taken part or no longer wish to take part. · Reassurances that their confidentiality will be maintained. · Thank participants for their time · Ask if there are any further questions

Answer 119

· The standardised instructions must give a step by step account of everything participants have to do from beginning to end and you need to consider the following things: · The task that participants have to do is made completely clear. · Include timings if possible so that participants know how long tasks are expected to take or how much time they have available. · Ask if there are any further questions before they start.

Answer 120

so ppt can understand fully what thry are doing avoid researcher accidentally changing the content of information they present from one ppt to another which could affect results replicability

Answer 121

a small–scale version of an investigation that takes place before the real investigation is conducted. The aim is to check that procedures, materials, measuring scales work and to allow the researcher to make changes or modification if necessary. Any problems that are identified are overcome before the full-scale experiment is carried out, thus increasing the validity of the results.

Answer 122

1. materials used 2. instructions clear 3. timing of experiment 4. any chance of demand characteristics 5. have variables been operationalised sufficiently

Answer 123

Are materials appropriate? E.g. are they clear enough for the participants to see/hear/read? Are they of the right level of difficulty? Is equipment working properly and positioned appropriately?

Answer 124

For example, a questionnaire can be piloted to make sure that the instructions and questions are clear and mean the same thing to all participants. Any problems can be rectified before the actual research begins. A questionnaire/interview should not have any ambiguous questions (questions that can be interpreted in more ways than one). There should not be too much jargon within your questions etc. An experimenter should also ensure that the instructions regarding the task are clear. Do the participants understand their role, how long the task will take etc?

Answer 125

is there enough time to complete the tasks or part of the tasks? Too much time or not enough time will affect the study

Answer 126

The participants are debriefed afterwards and asked about their experience of the study. This can reveal whether your study is truly valid as if they admit to guessing the aim, you will need to change the procedure e.g. turning an overt observation into a covert one, or in a questionnaire, including more filler questions.

Answer 127

Important for all studies but particularly observations that are using a behaviour schedule. For example, if you wished to observe aggressive behaviour in a playground, you could operationalise aggressive behaviour as the amount of times you observed hitting and pushing. A pilot study may highlight more items for your behaviour schedule that you didn’t think of e.g. biting. This is particularly relevant to observations where the behaviours must be measurable, cover ALL options and be mutually exclusive. Pilot studies are also a good chance to check inter-observer reliability and train observers in the use of the behavioural checklist.

Answer 128

Quantitative data is numerical. Quantitative data collection techniques usually gather numerical data in the form of individual scores. Data is open to being analysed and is easily converted into graph and charts.

Answer 129

Qualitative data is expressed in non-numerical formats and allows researchers to analyse feelings and thoughts that quantitative data cannot. It can take the form of written description or transcript from: interviews, diaries, and notes. It could also be in the form of pictures that participants may have drawn.

Answer 130

experiments observations - behavioural categories questionnaires - closed interviews- structured correlations

Answer 131

diaries case studies open ended questions unstructured interviews observation- written account

Answer 132

strength - objective -high reliability - precise numerical measures used limitation - lacks detail could miss important insights

Answer 133

strength rich and detailed- can gain new insights limitation -subjective- analysis open to interpretation - imprecise non- numerical measures used - low in reliability

Answer 134

This is sometimes called field research, and refers to the original data that has been collected specifically for the purpose of the investigation by the researcher. It is data that arrives first- hand from the participants themselves. Data which is gathered by conducting an experiment, questionnaire, interview or observation would be classed as primary data.

Answer 135

Strengths: It fits it purpose, in that it is authentic data obtained from the participants themselves for the purpose of a particular investigation. Questionnaires and interviews for example can be designed so that they target specifically the information the researcher requires. Limitations: It requires time and effort to gain primary data. For example planning an experiment can be time consuming, and costly to get resources, when compared to secondary data which can be accessed in minutes.

Answer 136

This is data that has been collected by someone other than the person carrying out the research. This means that it is data that already exists before the researcher has begun their investigation. This data is sometimes referred to as “desk research” and is often the case with secondary data has already been subjects to statistical testing and therefore significance is known. Secondary data may be located in journals, articles, books, or websites. Statistical information held by the government for example census records are examples of secondary data.

Answer 137

Strengths It can be inexpensive and easily accessed requiring minimal effort. When examining secondary data the researcher may find that the desired information already exists, therefore there is no need to conduct primary research. Limitations However, there may be a substantial variation in the quality, and accuracy of secondary data, information might at first appear to be valuable, but it may be out of date or incomplete. The research might not meet the researcher’s specific needs or objectives, so could affect the validity of their research.

Answer 138

Strengths: The Meta- analysis allows us to view data with much more confidence and results can be generalised across much larger populations. Limitations: However meta-analysis is much more prone to publication bias, sometimes referred to as the file drawer problem. The researcher may not reflect all the relevant studies, choosing to leave out the studies with no-significant results, leading to incorrect conclusions to be drawn as the results will be biased as not all data is represented.

Answer 139

This uses secondary data and refers to a process in which the data from a large number of studies, which have involved the same research questions and methods of research, are combined. The researcher may simply discuss the findings/conclusions which is qualitative analysis. They may additionally use a quantitative approach and perform a statistical analysis of the combined data. This may involve calculating the effect size; this is basically the dependent variable of a meta-analysis, and it gives an overall statistical measure of difference or relationship between variables across a number of studies.

Answer 140

1. measures of central tendency - single score representing the whole set of scores 2. a measure of dispersion - spread of the set of scores

Answer 141

mean median mode

Answer 142

Measures of central tendency are averages. These give you summaries of data which are easier to look at rather than lots of sets of data.

Answer 143

adding all the scores together and then dividing the number of scores

Answer 144

Strengths: Takes all the values of the scores into account, therefore is the most powerful of the measures of central tendency available. Limitations: Can be misleading if there are one or more extreme scores which are unrepresentative of the rest of the scores.

Answer 145

The middle score or measurement. To find the median scores are arranged in order (ranked) and the middle score is the median. If there is an even number of scores there will be two middle values. In these circumstances the two central values are added together and dividing by two.

Answer 146

Strengths: Unaffected by extreme scores – can be used with data from skewed distributions. Easier to calculate than the mean. Limitations: May be unreliable with small sets of data. It is not as sensitive as the mean, because not all scores are used in the calculation.

Answer 147

The mode is the most commonly occurring score in a data set.

Answer 148

Strengths: Shows the most important (frequent occurring) value in a set of scores, so the only one that can be used when data is in categories. Easy to find and not affected by extreme scores. Limitations: Does not work well with small sets of data or take into account exact values of each score. There may not be a single mode – scores can be bimodal if there are two modes.

Answer 149

refer to the spread of scores within a set of data the range standard deviation

Answer 150

This the simplest way to measure the dispersion of a set of scores. It is obtained by subtracting the lowest score from the highest score.

Answer 151

Strengths: Quick and easy to calculate compared to the standard deviation. Can be used with values that are on an ordinal (rating data) or interval (fixed intervals e.g. time) scale of measurement. Limitations: Does not take into account individual values, so does not use all the data collected. Seriously affected by extreme values in a given set of data as it is only taking the lowest and highest two scores into account.

Answer 152

This measures the average deviation (difference) of each score from the mean. Every score is involved in the calculation and it gives an indication of how far the average participant deviates from the mean. A small SD would indicate that most scores cluster around the mean score (similar scores), whereas, a large SD would suggest that there is a greater variance (or variety) in the scores.

Answer 153

Strengths: More sensitive dispersion method because all scores are used in the calculation, not just the highest and lowest scores. It is also not as easily distorted by one extreme score like the range. Limitations: Much more complicated to calculate than the range. As it requires the mean, it can only be used with interval data - because you wouldn’t use the mean for ordinal or nominal data. It is less meaningful if data is not normally distributed as the mean is used as the basis of the calculation, so if the mean is not representative of the data (has been affected by extreme scores), the SD won’t be either.

Answer 154

Data can also be presented in a table that you may be asked to interpret, or write a conclusion for. We don’t present raw scores in a table but instead we represent descriptive statistic

Answer 155

Measurements of participants’ behaviour obtained directly from a research study are called raw scores. They have not yet been organised and processed. It is often necessary to rearrange the raw scores into a frequency distribution. This is a table showing the frequency of all the scores. The total number of times a behaviour occurs is called a frequency. The table on the left shows the raw data for an investigation into the amount of hours male and female students spent revising the day before an exam. Turn this raw data (left) into a frequency distribution table (right).

Answer 156

Graphs and charts act as visual aids, which help us make sense of data obtained from psychological studies. They aim to provide an overall picture, which helps to show patterns and trends.

Answer 157

Consist of a series of vertical bars of equal width, which can be used to illustrate the mean/median of different types of data. They provide a simple and effective way of comparing data. It is usual to draw each bar separated from each other because each bar represents a different category of data (the example shows the effectiveness of two types of police interview).

Answer 158

measuring the relationship between two co variables

Answer 159

Pie charts are used to show the frequency of categories as percentages. The pie is split into sections, each one of which represents the frequency of category. The sections are colour coded, with an indication of what each section represents.

Answer 160

Line graphs represent continuous data and use points connected by lines to show how something changes in value, for instance, over time. Typically, the IV is plotted on the X-axis and the DV on the y-axis. For instance, in an investigation of how the passage of time affects our ability to remember information, the decline in recall would be shown as a continuous table.

Answer 161

These are mainly used to present frequency distributions of interval level data (data on a scale). The horizontal X axis is on a continuous scale, e.g. time in seconds. The Y axis displays the frequency that the particular score occurred. On a histogram, there are no spaces between bars, because the bars are not considered separate categories. Sometimes though, the scale is condensed into number ranges to reduce the number of bars involved e.g. 1 bar = scores 1-10. Histograms are then used to check the distribution of the data (whether it is normal or skewed)

Answer 162

normal positively skewed negatively skewed

Answer 163

is a symmetrical representation of frequency data that forms a bell- shaped pattern. Most people are scoring somewhere on the middle of the scale, with few extreme low or high scores. The mean, median and mode are all located at the highest peak.

Answer 164

Skewed distributions are where scores are not equally spread around the mean. In these distributions, you normally have data with many more values at one end of the scale than the other. A number of outliers (extreme scores) create a ‘tail’ to the distribution, and influence the mean away from where most people scored.

Answer 165

the highest point, represents the most common

Answer 166

this is a type of distribution in which the long tail is on the positive (right) side of the peak and most of the distribution is concerned on the left. In the case of a positive skew, most people are scoring lower on the scale, but a few extreme high scores are influencing the mean to be higher than the majority of scores in the data. This can sometimes be caused by a measure (e.g. a test) that is too difficult.

Answer 167

A type of distribution in which the long tail is on the negative (left) side of the peak and most of the distribution is concentrated on the right. In the case of a negative skew, most people are scoring higher on the scale, but a few extreme low scores are influencing the mean to be lower. This can sometimes be caused by a measure (e.g. a test) that is too easy.

Answer 168

Content analysis is used to quantify qualitative data and is commonly performed with media research. In this method, the researcher analyses written, verbal and visual communication numerically. The researcher identifies relevant categories and then counts how often they occur. This thereby converts the raw qualitative data into quantitative frequencies, i.e. a count is made of the number of times certain selected types of event happen.

Answer 169

content analysis thematic analysis

Answer 170

1) A sample of materials is examined by at least 2 separate researchers. 2) Reading / looking over the materials leads to the identification of suitable relevant categories. 3) The categories are then agreed on and clearly defined / operationalised 4) ‘Coding’ then occurs by two different researchers separately reading back over the material and counting the ‘frequency’ with which each category occurs 5) The frequencies tallied by each researcher are then compared and checked for inter-rater reliability

Answer 171

Thematic analysis is a qualitative method which identifies, analyses and reports themes (patterns) within data. The identified themes then become the categories for analysis. Thematic analysis goes beyond just counting words or phrases, and involves identifying ideas within data. Analysis can involve the comparison of themes or the identification of co-occurrences of themes. The key difference between content analysis and thematic analysis is that the latter remains qualitative; the write up involves a detailed description of the themes found.

Answer 172

1) The data is read through by at least 2 researchers repeatedly to identify themes - Themes refer to any idea (implicit or explicit) that is recurrent in the qualitative text being analysed (i.e. crops up repeatedly) 2) The data is read and re-read until all the themes have been identified. 3) The themes arrived at between the researchers are then compared to see if similar ideas have been arrived at. 4) The themes may then be developed into broader categories. 5) The themes can then be applied to a new set of data to check for validity. Assuming that the themes identified fit the new data then they can be said to be valid. 6) The report is then written up using examples of quotes that best represent the themes.

Answer 173

strengths It is a useful tool for gathering data from a wide range of areas, from children’s drawings to the aggressive content of books and films. Furthermore, any data gathered has high ecological validity as the participant is talking about their feelings and real life experiences. It can get around many of the ethical issues normally associated with psychological research. Much of the material than an analyst might want to study, such as TV, adverts, films, personal adverts etc. may already exist within the public domain. Thus, there are no issues with obtaining permission. Reliability of Content Analysis: Reliability is easy to establish in content analysis because it can be easily replicated using the same coding units and qualitative materials. Limitations of Qualitative Analysis (Content and Thematic): It is sometimes difficult to arrive at objectively operationalised coding units, and can be time consuming to carry out. It is prone to subjective interpretation – in qualitative analysis, people tend to be studied indirectly, so the communication they produce is usually analysed outside of the context within which it occurred. Therefore, there is a danger that the researcher may attribute opinions and motivations to the speaker or writer that were not intended originally. Though this is more an issue with thematic analysis than content analysis.

Answer 174

Peer review is the process of subjecting a piece of research to independent scrutiny by other psychologists working in a similar field, before it can be published, who consider the research in terms of its validity, significance and originality prior to publication. Peer reviewed research may be accepted, sent back for revisions or rejected.

Answer 175

Several expert reviewers are sent copies of the researchers work by a journal editor. They read the manuscript and assess it carefully highlighting weaknesses or problem areas. Then they send it back to the editor with comments, including possible changes that may need to be made. The peer reviewers also recommend whether they think the paper is suitable for publishing or not.

Answer 176

.It is difficult for authors and researchers to spot every mistake in a piece of work. Showing the work to others increases the probability that weaknesses will be identified and addressed. · It helps to prevent the dissemination of irrelevant findings, unwarranted claims, unacceptable interpretations, personal views and deliberate fraud. · Peer reviewers also judge the quality and the significance of the research in a wider context. · This process ensures that published research can be taken seriously because it has been independently scrutinised by fellow researchers.

Answer 177

· Peer reviewers may be biased and lack objectivity: The reviewers’ theoretical viewpoint may differ from that of the researcher, (e.g. would Freud have recommended that a journal published work by Skinner?); There may be gender bias - the tendency to favour male researchers; Institution bias - the tendency to publish research from high status institutions. Funding bias - Research may only be published due to their funding, they may only want certain research to be seen as scientifically acceptable. Difficulty of anonymity - reviewers may criticise rivals work. · File drawer phenomenon may occur. This may happen as peer reviewers tend to favour positive results over negative ones. Negative findings are either not published or ‘left in the researcher’s file drawer’. As a result, there is only a partial understanding of the topic. · Preserving the status quo. Most research builds on previous knowledge and research which does not agree with previous work may be rejected. The peer review process may keep things the same and prevent new knowledge coming through. · Values in science. Although psychologists aim to be objective, many philosophers of science suggest it is impossible to separate research from cultural, political or personal values. If the author and the reviewer share these values they may be published as objective, in fact when they are not.

Answer 178

One of the wider concerns for psychology as well as science in general is what the implications of research are for the economy. By implications, we mean what have we learned from the findings of psychological research? How does this influence (affect, benefit or devalue) our economic prosperity? Psychology creates practical applications in everyday life for the better of society. Therefore research contributes to the economy in a substantial way.

Answer 179

Bowlby’s research on attachment suggested that a child can only have a secure and lasting monotrophic attachment with its mother. At the time it was seen as the mother’s role to care for the child and the father to provide economic stability. However, more recent research has found that children make multiple attachments, most notable with the father; this is deemed as necessary for healthy psychological development to have an attachment and bond with both parents. The knowledge that both parents are capable of providing emotional support for healthy psychological development, has the implication for the economy by which that it allows for flexible working households. In households where the mother is the higher wage earner and so works longer hours, many couples share child care responsibilities, this mean that modern parents are better equipped to maximise income and contribute more to the economy.

Answer 180

Absence from work cost the economy an estimated £15 billion a year. A recent government report revealed that a third of all absences are caused by mild to moderate mental health disorder such as depression, anxiety and stress. A diagnosis and access to treatment of a mental disorder such as psychotherapeutic drugs or CBT can help to provide management of symptoms and assist in return to work. Therefore, the economic implication is that it is vital to research the causes, and treatments of mental illness to support a healthy workforce.

Answer 181

Reliability refers to consistency, i.e. the ability to get the same results. If a study is repeated using the same method, design and measurements, and the same results are obtained, the results are said to be reliable.

Answer 182

internal external

Answer 183

This concerns the extent to which a particular measure used in an investigation (e.g. a questionnaire or test administered) is consistent within itself. For example, an IQ test should be consistently measuring intelligence throughout all the questions on the test, and if this is the case, it has internal reliability.

Answer 184

This is more about reliability over time and concerns the extent to which the results of a measure are consistent from one use to another. For example, if somebody rated as aggressive on a questionnaire was to repeat that test a week later, they should get a similar score if the test has external reliability. Similarly, if you repeated a whole investigation, you would expect the results to be the same for those results to have reliability.

Answer 185

1. split half method 2. test re test 3. inter observer reliability

Answer 186

The split-half method measures internal reliability by splitting a test into two (e.g. splitting a questionnaire into odd questions and even questions) and having the same participant complete both halves. If the two halves of the test provide a similar score, this indicates that the test itself has internal reliability. You would compare the scores from the two halves using a correlational analysis – where a significant positive correlation of 0.8 or higher indicates reliability.

Answer 187

Test-retest is used to measure external reliability, by giving the same test to the same participants on two different occasions. Again, scores from the 2 tasks (the test and the retest) are compared using correlational analysis. A significant positive correlation (of 0.8) indicates reliability.

Answer 188

This is used for observations and sometimes interviews. It is a means of assessing whether different observers are viewing/rating behaviour in the same way. Two observers would carry out the test separately and then the observers’ sores would be analysed using a correlation. A high positive correlation (0.8) between the two sets of observers scores would indicate that they are categorising behaviour consistently. Inter-observer reliability in particular can be improved by developing clearly defined and separate categories of observational criteria (operationalisation) and training researchers on the use of the test to make sure it is being applied consistently – this can be done as a pilot study.

Answer 189

· Improving the reliability of experiments can be done by standardising the testing procedures. Tighter controls over EV’s means that the procedure is more tightly standardised, so that when it is repeated it is conducted in the same manner. It is also important to make sure the researcher conducting the study is the same too, which is particularly important with interviews where differences in interviewers style of questioning could affect the answers participants give. · Properly operationalising variables will improve reliability, as this ensures that another person repeating the test will be manipulating the IV and measuring the DV in the same way. For observations, this involves using behavioural categories which are properly constructed (categories not overlapping etc.) to ensure recording of behaviour is consistent. · A pilot study can be used to check the proposed method of measurement works properly and that participants can use the apparatus successfully. For observations, a pilot study will involve training observers in the use of the behavioural checklist to improve inter-rater reliability. A pilot study may also be important for interviews to check that, if different interviewers need to be used, they are interviewing participants in the same way (e.g. one isn’t asking particularly confusing questions).

Answer 190

Validity concerns accuracy; the degree to which something measures what it intends to. Therefore, validity refers to how accurately a study investigates what it claims to. Validity also refers to whether the findings are accurate beyond laboratory research settings, or in other cultures and time periods (can be generalised).

Answer 191

internal external

Answer 192

This concerns whether the research is accurate in itself, and whether the researcher has measured what they intended to e.g. whether the researcher has definitely measured the effect of the IV on the DV and not accidentally measured the effect of an extraneous variable instead, such as demand characteristics.

Answer 193

This concerns whether the results are still accurate in other settings (ecological validity), other people (population validity) and over time (temporal validity). In other words, external validity is about whether research can be generalised to other situations.

Answer 194

Face Validity The extent to which a measure, at ‘face value’, looks like it is measuring what it intends to. E.g. A test of IQ which only requires participants to memorise items doesn’t look on the surface to be a valid measure of IQ. Concurrent Validity Involves correlating scores on a new test of unknown validity with another existing test that is known to be valid and trusted. For example, a researcher gives the same participant a new personality questionnaire and at the same time, a well-known, valid personality questionnaire. If both questionnaires give similar results, it is likely this new questionnaire has high validity.

Answer 195

Meta-analysis This involves the comparison of findings from many studies that have investigated the same hypothesis. These studies are often carried out in different time periods, in different countries and may have been carried out using different methods (e.g. laboratory vs field experiment). Findings that are consistent across populations, locations, and periods in time in the meta-analysis indicate external validity e.g. Van Ijzendoorn and Kroonenberg’s (1988) investigation of the strange situation.

Answer 196

Problems with internal validity may be caused by confounding variables, demand characteristics and investigator effects. Ways of improving internal validity include: Choose the appropriate experimental design · Control extraneous variables § Single blind technique § Double blind technique

Answer 197

Sampling – selecting a sampling method that will enable the investigator to collect a representative sample (typical of the whole population), e.g stratified sampling. Repeat the study on a different sample – if the findings do represent other relevant populations then you can say that the findings have population validity. Realism – conduct the investigation in the most realistic setting possible, ensuring if in a laboratory that the setting is as true to life as possible.

Answer 198

T-heory construction and revision H-ypothesis Testing E-mpirical Methods P-aradigm and paradigm shifts R-eplicability O-bjectivity F-alsification

Answer 199

Theory construction is a major feature of scientific disciplines. A theory is a collection of general principles which can explain a particular event or behaviour. For example, Social Learning Theory can explain how young children acquire behaviours like aggression. Popper saw the scientific process as starting with very tentative theories based upon observation (‘hunches’), which are then used to generate hypotheses (testable predictions). These predictions are investigated using rigorous techniques (see ‘empirical methods’) and analysed to see whether they are either supported or rejected. These investigations yield more information about the world, which is used to develop the theory; if the original tentative theory cannot account for this new information, it leads to adjustment of the theory (‘theory revision’). The scientific cycle then begins again with this revised theory. Through this cycle, the theory should be gradually gaining closeness to the truth.

Answer 200

An important part of theory construction and revision is hypothesis testing. Theories should be able to propose several clear hypotheses which can then be subjected to rigorous testing. A hypothesis that is consistently supported leads to the development of a theory, whereas rejection of the hypothesis would lead to theory revision. The process of deriving new hypotheses from an existing theory is called deduction. For example, our hypothesis based on social learning theory may be that children with parents rated as more aggressive will themselves display more aggressive acts towards a bobo doll than those who do not have aggressive parents. If we find this to be the case, it strengthens social learning theory as an explanation.

Answer 201

Science uses empirical methods. Empiricism refers to the view that gathering data and evidence from experience (sensory information) is central to the scientific method, rather than simply relying upon our own viewpoints, which can be influenced by our desires. This experience is best gained through controlled experiments or observation. Laboratory experiments are considered the most empirically based research method because cause and effect can be established. By using empirical methods, it also ensures that researchers are basing their conclusions on experience rather than viewpoint, which makes it more objective and therefore more scientific (see objectivity).

Answer 202

Kuhn (1962) argues that what separates sciences from non-sciences is a paradigm - a set of shared assumptions and agreed methods within a scientific discipline. For example, Biology is characterised as having a set of assumptions at its core e.g. role of genetics, evolution. Kuhn argued that Popper’s idea of the scientific method (see theory construction) isn’t how science works. Rather than slow modification of theories over time until they become true, Kuhn argues that progress within a science occurs due to a scientific revolution, when there is too much contradictory evidence to ignore. These are called ‘paradigm shifts’ and it is thought often come initially from a minority position. For example, the shift from Newtonian paradigm in physics to Einstein’s theory of relativity. Paradigm and paradigm shift example At one point in history, scientists were convinced that the earth was flat and that if we kept going, we would eventually fall off. This was the accepted paradigm by nearly everyone living during this time. However, a minority of researchers questioned this, and perhaps explorers travelled the word and provided evidence that they did not ‘fall off’, and this contradictory evidence became hard to ignore. This was then accepted by others and a paradigm shift had occurred. Another example in Psychology is the paradigm shift from behaviourism, the accepted paradigm of how to study human behaviour at the time, to the cognitive approach which considered the thought processes between sensory input and behaviour. However, Kuhn has suggested that social sciences (including psychology) lack a universally accepted paradigm because there are too many conflicting approaches, therefore cannot qualify as a true science.

Answer 203

Another major feature of a science is replicability. If a scientific theory is to be truly validated, experiments must be repeatable over different contexts and circumstances and findings obtained must be consistent. The more evidence in support of a theory across different times, populations and settings, the stronger the theory becomes. Therefore, if research cannot be repeated, it is not scientific. Replicability also has an important role to play in checking the validity and reliability of research. It is through replication that any inaccurate findings, which may have occurred due to mistakes in the research procedure (or simply by chance!) are discovered. Results that are consistently repeated are more likely to be reliable. In order for replicability to be possible, researchers must report their investigations with as much rigour and precision as possible, so that other researchers can verify their work and the findings they have obtained

Answer 204

For research to be scientific, researchers must maintain objectivity as part of their investigations, and keep a critical distance by basing their research on direct experiences and not biased viewpoints (empiricism). They must not let their personal opinions discolour their research or influence the behaviour of participants in order for it to be scientific. Typically methods most associated with greater control e.g. laboratory experiments, tend to be most objective. To lessen the possibility of bias, researchers also aim to use standardised instructions, operationalisation, the double blind technique etc. Replicability and peer review are two ways of checking that research is objective.

Answer 205

Popper suggested that genuine scientific theories should hold themselves up for hypothesis testing and the possibility of being proven false. He believed that even when a scientific principle had been successfully and repeatedly tested (even over many occasions!), this was not enough to prove it to be true. Instead, it had simply not been proven false - yet. This became known as the theory of falsification. Popper drew a clear line between good science – where theories are constantly challenged, and what he called ‘pseudosciences’ which couldn’t tested or be falsified. Those theories that survive most attempts to falsify them through repeated replications, become the strongest – not because they are necessarily true – but because, despite the best efforts of reseachers they have not been proved false. Hence why psychologists avoid using phrases such as ‘this proves’ in favour of ‘this supports’ or ‘this seems to suggest’ and provide two hypotheses for their research - null and alternate. Replication of research where researchers consistently fail to falsify is the accepted way of validating a theory.

Answer 206

1. title 2. abstract 3. introduction 4. method 5. results 6. discussion 7. references 8. appendices

Answer 207

To tell the reader what the report is about – concise but informative.

Answer 208

A brief summary (150 – 200 words) of the key points of the study that appears at the start of the report. This helps the reader to decide whether the full report is worth reading, a bit like the blurb of a book. The abstract usually includes a sentence on the aim, research question, the method used, findings and conclusion.pu

Answer 209

Background to the research area and rationale (why the study was conducted). The background will include a literature review of relevant past studies and theories, which are used to set the current research paper in the context of relevant past research and why the current research has been conducted (e.g. to address a ‘gap’ in the research. The introduction ends by specifying the research aims and predictions/hypotheses for the study. A ‘funnel’ technique is used, starting from broad theories and past research and steadily narrowing down to the specific study in question.

Answer 210

Describes how the study was carried out in sufficient detail for someone else to be able to replicate it. This enables checks for internal and external validity, confounding variables, demand characteristics, sampling errors etc. (Disagreements between psychologists are often about the method used.) The method has four sub-sections: Design: an outline of the method used e.g. laboratory experiment, including the type of design e.g. repeated measures, number of groups or conditions, details of the variables (IV and DV) being investigated and operationalised. Any controls needed. Participants: Information on the sample of participants used including number, gender, age and the sampling technique used. Apparatus/Materials: The materials used to carry out the research, e.g. questionnaire, stimulus material, standardised instructions of what the participant is to do, information sheet with the consent form. Important!! You may be asked to write standardised instructions in the exam. You must write them as though you would be reading them out formally to the participant. Procedure: This details how the study was carried out, step by step, including how the participants were allocated, instructions used, how the data was collected etc.

Answer 211

Summarises the findings of the research clearly and accurately. There is normally a section on descriptive statistics and also inferential statistics. Descriptive statistics: Measures of central tendency (such as the mean) and dispersion (like the standard deviation) are shown clearly in labelled tables and graphs. Tables and graphs should be clearly titled and labelled, and units of measurement specified. Tables should be numbered and titled above the table, figures and graphs below. Inferential statistics: Reasons for selecting a particular statistical test are given, as well as what it tests for. Data on this is also shown (calculated and critical values, significance level and whether it was one or two tailed) followed by analysis of whether the null hypothesis can be rejected or accepted.

Answer 212

his section explains what the results mean and is broken down into several sections. Explanation of the findings: Key findings are described that relate to the aims and hypotheses. All findings should be presented (including unexpected and contradictory ones), plus an explanation of what the findings show and a possible explanation for why they occurred. Relationship to background research: Research is presented and discussed in terms of previous research findings presented in the introduction. Aspects of the design that may account for differences in the findings from previous studies are outlined. Limitations and modifications: Possible sources of error in the study, like flawed measurement techniques, poor sampling, lack of controls and/or poor procedures are outlined and discussed. Possible ways of rectifying these faults are suggested. Implications and suggestions for future research: Here, further research studies are suggested based on the findings of the current study. For example, maybe the study did not look at a particular sample of people where results could have been different, or suggest conducting the study in a different setting with more validity. Also presented here are any implications and applications from the findings of the study.

Answer 213

Information on sources of information used in the report shown in alphabetical order. This provides evidence for the reader. e.g. Miller, G.A. (1956). The magical number seven plus or minus two: some limits on our capacity for processing information. Psychological review, 63, 81-97.

Answer 214

nominal ordinal interval

Answer 215

This involves counting frequency data in categories. For example, how many people have a dog, a cat, or a guinea pig etc. as a pet?. Tally charts are used to record this type of data. Nominal data has weaknesses, as it is the crudest, most uninformative type of data. For example, it doesn’t tell us how much they prefer having a dog over other types of pets. The mode is used to analyse this data as it is recording the most frequently occurring score.

Answer 216

This is when data is ranked so that it is possible to see the order of scores in relation to one another (e.g. 1st, 2nd, 3rd or ‘most to least attractive’). Ordinal data also uses rating scales, but these are arbitrary (not precise, may mean different things to different people) e.g. rate how happy you are on a scale 1-10. The median is used to analyse this type of data. A strength of ordinal data is that it is more informative than nominal data (scores are now being put into some order). However, it still lacks information, making it less sophisticated than ordinal. For example, the finishing places in a race may show which athletes are faster than others by ranking them, but it doesn’t inform us about how much faster 1st was from 2nd and distances between the ranks may not be equal - it may be that only 1 second separated the finishing times of 1st and 2nd place athletes, but 50 seconds between 2nd and 3rd! Also, arbitrary scales are subjective (one persons 7 is different to anothers)

Answer 217

This is a more sophisticated level of data. It not only gives the rank order of scores but it also details the precise intervals between scores. It uses precise, standardised units i.e. minutes, degrees, miles. A strength of this type of data is that it is the most informative and also accurate form of measurement as the distances between the units are all equal and standardised. For example, one second in time is the same length as any other second on the measurement. The mean is used to analyse this type of data.

Answer 218

to discover if your results are significant or not A result is significant if there is a low probability that the result has occurred due to chance factors i.e. we have found a real difference or a real relationship. we need to examine the probability that our results occurred due to chance, which (ideally!) should be as low as possible. Inferential statistical tests calculate this probability and give it to us as a ‘p value’ (p standing for probability) expressed a decimal point. E.g. if the statistical test reveals that P=0.05, this means there is a 5% likelihood that our results occurred due to chance. P=0.1 means there’s a 10% likelihood that results were due to chance.

Answer 219

This means that a result is significant if there is less than 5% probability that results are due to chance, and therefore researchers can be 95% confident that they have found a real result.

Answer 220

When a researcher first starts their investigation, they form two contradictory hypotheses: · The experimental/alternative hypothesis which says “There is a significant difference/relationship….” · and the Null which says: “There is NO a significant difference/relationship and that any observed differences are due to chance” with the aim of the research being to try and accept the experimental/alternative and reject the null hypothesis. So through conducting inferential statistics on their results, the researcher is essentially determining the likelihood that the null hypothesis is true, and whether they can then reject it or not. Based on the p<0.05 significance level, this means that: · If there is more than 5% chance that the null hypothesis is true (e.g. P=0.07) then we must accept it. · If there is less than 5% chance that the null is true (e.g. p=0.03) then we can reject it and accept our alternative hypothesis.

Answer 221

1. Whether you are looking at a difference or relationship 2. The type of data you have: Nominal, Ordinal or Interval/Ratio 3. Whether you have carried out a repeated measures design or independent measures design. 4. For some, whether the data is normally distributed.

Answer 222

chi squared (x2)

Answer 223

man whitney U (U )

Answer 224

unrelated t test (r)

Answer 225

sign test (S)

Answer 226

wilcoxen test (t)

Answer 227

Related t test (R)

Answer 228

chi squared (x2)

Answer 229

spearmans roe (Rs)

Answer 230

Pearsons product moment (R)

Answer 231

Determining whether a result is significant or not involves comparing a calculated value to a critical value.

Answer 232

- The calculated value (sometimes referred to as the observed value) is what the result of the statistical test gives you ie. what you’ve calculated. (The calculated value is always reported after the short letter for the test e.g. for a sign test it’s S=, or Spearman’s rho is Rs=)

Answer 233

- The critical value is a number that you look up in a table, which will be particular to the test you are using.

Answer 234

1. State whether the calculated value (S, U, T, Rs, or X2) is greater than/less than the critical value. Make sure to put the values of each in brackets. 2. State whether the results are significant or not. 3. Then state whether the null is accepted or rejected, then the experimental (always deal with the null first). 4. Write out the relevant hypothesis (whichever you have accepted). 5. Then lastly report the figures in brackets. a. Firstly, report again the calculated value. b. Then the number of participants who were in the analysis, written as N=. (or N1= and N2= if it is a Mann Whitney, or df if it is Chi Squared). c. Finally report whether p was more (>) or less (<) than 0.05. If your experimental hypothesis is accepted, write p<0.05. If your null hypothesis has to be accepted, p is higher than 0.05 and therefore should be written like p>0.05. d. Then write whether the hypothesis was one or two tailed.

Answer 235

Firstly, we need to convert the results to nominal data before we can do the sign test on it. We do this by working out which participants became more talkative (had a higher word count score) after the energy drink and which became less talkative (produced a lower word count). The categories therefore being MORE (+ sign of difference) or LESS (- sign of difference). 2. Then, we need to total up the minuses and the pluses. Those with no differences are discounted. - = 3 + = 6 3. The calculated value in a sign test is simply the number of the least commonly occurring sign, in this case the minuses. We call this calculated value ‘S’ in a sign test, therefore in this worked example S=3. If the number of minuses and pluses was equal (e.g. both 3), the calculated value would still be 3.

Answer 236

Psychologists believe that this margin of error is too high and we are thus likely to make what’s called a Type I error; this when we wrongfully say that a result is significant and reject the null hypothesis (when really if we’d have used a stricter level of significance e.g. p<0.05, it would have been non-significant).

Answer 237

Conversely we might be considering the effectiveness of a new drug treatment for depression and so we set our significance level quite high in the hope of reducing our margin of error e.g. p<0.001. With this level of significance, what margin of error (likelihood of chance) are willing to accept? ________. Using this level of significance is so strict that it only allows for the tiniest margin of error. However, in using this level of significance we may wrongfully say a result is non-significant and accept the null hypothesis when actually (if we’d have used p<0.05) we have found a significant result. Psychologists use the p<0.05 because they believe it strikes the best balance between Type I and type II.

research methods completed Flashcards

booklet number 1 and 2 finished