Methodology Flashcards

Question

What is a measurement question?

Answer 1

Questions that try to find out how to measure variables, e.g. how do we measure psychopathy?

Answer 2

These provide a basic picture of the extent and scope of a phenomenon. They try to describe things. e.g. How can a criminal career be characterised?

Answer 3

These go beyond descriptions, and begin to probe for details surrounding the phenomenon. They often involve conducting in-depth interviews of people affected by some phenomenon/relationships between phenomena (but not why there is a relationship- that is explanatory). Studies may also attempt to narrow possibilities of causes of phenomena- but they do not attempt to identify the causes. E.g. does marriage reduce crime?

Answer 4

These seek to identify one or more factors that actually bring about a phenomenon. What mechanisms are there that explain why a specific relationship exists? How does X happen etc.? e.g. Why do people get involved in crime?

Answer 5

After one or more causes of a phenomenon have been identified, scientists sometimes seek to alter the prevalence of that phenomenon. When this is done, research designed to assess how well the intervention worked is needed. E.g. what is the effect of a custodial sentence on reoffending?

Answer 6

- Clear - Focused - Relevant - Testable

Answer 7

Empirical phenomena that take on different values or intensities. The measurable representations of the constructs that you are interested in. Typically, these variables are things that we would turn into some kind of data file. If we want to use variable in statistical tests, we have to assign numbers to words.

Answer 8

Any phenomenon that always takes on the same value or intensity.

Answer 9

A given individual will usually be more criminal at some times in their life than others (usually a a gradual curve, peaking in the middle)

Answer 10

Essentially what you find in the dictionary. Familiar terms are used to describe a word or phrase.

Answer 11

These specify in empirical terms precisely what should be done to observe variations in the variable being studied, or exactly how to measure the variable. The process of making an abstract concept (the construct) measurable, as a variable.

Answer 12

An abstract concept, undefined. You can apply operational and conceptual definitions to it.

Answer 13

Conceptual: An act or omission that violates a law, and is punishable by the State, typically involving harm to others, society or property. Operational: Any behaviour that is formally labelled as illegal by legislative statutes (leading to arrest, prosecution, and punishment)

Answer 14

Conceptual: A person who suffers harm, injury or loss due to another's actions. Operational: An individual recognised by the criminal justice system as having been harmed by a crime.

Answer 15

Conceptual: The number of crimes committed in a particular area, divided by the the number of inhabitants in that area Operational: Find the number of crimes committed over the relevant time period (official data and/or self report), and divide this number by the total number of inhabitants. Then multiply by 100,000 for the rate per 100,000 people.

Answer 16

1. Demographic variables 2. Cultural variables 3. Social institutional variables 4. Behavioural and personality variables 5. Cognitive variables

Answer 17

These pertain to basic human characteristics like age, gender, ethnicity. They are widely used to describe human populations.

Answer 18

Within this category, there are 2 subcategories: - Artificial cultural variables: The wide variety of material things characterising most human societies, such as the homes they occupy, the neighbourhoods they live in etc. - Customary cultural variables: Values and practices that most members of a particular society share, e.g. languages, day-to-day customs we perform and expect from others, rules we follow.

Answer 19

The wide variety of material things characterising most human societies, such as the homes they occupy, the neighbourhoods they live in etc.

Answer 20

Values and practices that most members of a particular society share, e.g. languages, day-to-day customs we perform and expect from others, rules we follow.

Answer 21

These have to do with the functioning of social institutions and tangible manifestations of culture. Economic variables, like national unemployment rates, inflation rates, homicide or assault rates, as well as neighbourhood (dis)organisation.

Answer 22

How we act as individuals. It is difficult to make a clear distinction between the two, but behavioural pertains to more specific acts, while the personality pertains to acts that are much more consistent and characteristic of a particular person.

Answer 23

These encompass the emotions we feel, the individual attitudes we hold, the thoughts and intellectual abilities we possess, and the mental health or illnesses we have.

Answer 24

1. Self-report 2. Archival data 3. Observation

Answer 25

A data collection method. Asking a person questions about themselves, and receiving answers which are then studied

Answer 26

1. Interviews 2. Questionnaires

Answer 27

A type of self report, which is a type of data collection method. Interviews involve face-to-face questions, generally qualitative data, but there can be exceptions (quantitative)

Answer 28

A type of self report, which is a type of data collection method. Questions given to respondents in written form- can be seen (often) as quantitative data, since there are ranking systems, but there can be exceptions (qualitative)

Answer 29

A type of self-report (questionnaire), which is a type of data collection method. They involve asking large numbers of people if they have been criminally victimised within some specific time frame, regardless of whether they reported the incident to police.

Answer 30

- Can collect information known only to respondents - Researcher has control over the data collection - Includes unreported crimes - If re-taken, trends can be tracked/predicted - Anonymity of online surveys can prompt people to give honest answers

Answer 31

- "victimless" crimes are harder to include, e.g. drug dealing, gambling, and even victims of murder - Often surveys are given to households, thus excluding crimes against commercial establishments - No guarantee that information given by the respondents is true/valid - Concerns over memory and misreporting - Can be time-consuming - Many surveys are convenience based, and the respondents are typically students

Answer 32

The proportion of people who are (committing crimes)

Answer 33

The number of (criminal) events that occur.

Answer 34

People's varying propensity to commit crimes and other antisocial acts.

Answer 35

A type of recall bias. The temporal displacement of a past event in someone's mind, specifically: - Recent events are perceived as more distant than they are - Distant events are perceived as more recent than they are

Answer 36

The tendency of respondents to give answers they believe to be more socially acceptable or favourable.

Answer 37

1. Make a list of things you want to know 2. Draft different parts of the questionnaire 3. Write the items and response options 4. Check the items and response options 5. Add instructions/notes for the participants

Answer 38

A question or statement in a questionnaire, which you want some kind of response to.

Answer 39

1. Descriptive items 2. Recall items 3. Attitudinal items 4. Projective items

Answer 40

A question or statement in a questionnaire, that ascertains descriptions about a person, e.g. what is your age? or what is your marital status?

Answer 41

A question or statement in a questionnaire, that asks someone to recall something from the past, e.g. have you ever been a victim of X?

Answer 42

A question or statement in a questionnaire, that asks for attitudes or opinions from the respondent, e.g. what do you think of the death penalty?

Answer 43

A question or statement in a questionnaire, that asks about the future. They can sometimes be called scenarios or vignettes. E.g. What would you do if you witnessed a crime? or are you planning on buying a house in the next year?

Answer 44

Also known as scales. Unlike simple constructs that can be assessed with one question, e.g. age or sex. We should have some kind of measurement, or instrument, to measure them. We can use a measurement question with multi-item questionnaires for this. E.g. If things I do upset people, it's their problem, not mine.

Answer 45

A multi-item questionnaire.

Answer 46

Any physical thing used to collect data.

Answer 47

The range of answers that researchers offer subjects in connection with the questions that are asked.

Answer 48

The process of deciding how data will be entered into a standard format for analysis.

Answer 49

1. Open-ended 2. Fill-in-the-blank 3. End-anchored continuum 4. All-point anchored 5. Ranking

Answer 50

They are response options for items. They present a question/statement without providing constraints on how they are answered. E.g. how would you feel if you ended up in prison?

Answer 51

They are response options for items. Respondents must write/say a word or phrase missing from a statement. E.g. I consider __ the most serious crime.

Answer 52

They are response options for items. Subjects are allowed to respond along a continuum with extreme opposites at each end. Demarcated linear forms are from adjective - adjective. (e.g. Strong - weak, or good - bad) Numerical forms are from 1 - 10.

Answer 53

They are response options for items. Options range from strongly agree to strongly disagree, and are also known as Likert scales.

Answer 54

They are response options for items. Respondents are given a list to rank in some way.

Answer 55

- The amount of coding to do later - Try to measure variables at an interval level or higher. - Do not switch too much between types of response options. - Give subjects considerable latitude in their responses Items should be clear, neutral, relevant, should provide useful response items, and ask for one thing at a time.

Answer 56

- Rhetorical and leading questions - Conjunctive items (asking multiple things at once) - Yea-saying items (philosophical sounding questions that tend to compel respondents to agree) - kernel-of-truth items (partially true statements, especially ones which can sometimes insult) - Questions about behaviour not bounded by time (and sometimes place)

Answer 57

A variable that exists in two or more segments, regardless of how researchers decide to measure it. E.g. Sex, religion, nationality. Normally, discrete variables are measured at the nominal or ordinal level.

Answer 58

Does not exist in segments (as in discrete), but instead varies gradually from low to high, weak to strong etc. E.g. Age, academic ability, fear of crime. Normally, continuous variables are measured at the interval or ratio level.

Answer 59

These break down a major question into several components, and then focus on answering each component.

Answer 60

When a journal/manuscript is sent to educated peers for review before publishing.

Answer 61

When a journal/manuscript is sent to educated peers for review before publishing, and the reviewers (also known as referees) are not known to the author.

Answer 62

A data collection method. Official data, looking back in time, always collecting empirical data, but not always collected with a specific research project in mind. E.g. police records

Answer 63

The analysis of archival data

Answer 64

Relatively recent data, usually since the turn of the 20th century, often compiled by agencies that still exist. E.g. police data (modern)

Answer 65

May be centuries old, often (usually) collected by agencies that no longer exist. E.g. church or jail records.

Answer 66

Archival data that is open to the public.

Answer 67

Archival data that needs permission to access it, often because the data is sensitive.

Answer 68

- Cost and time efficient - Often contains more information than researchers could collect alone

Answer 69

- Can be outdated - Data is often gathered at a group level, leading to ecological fallacies - Differences between countries (e.g. resources, definitions, biases) - Not all data is gathered for research purposes, so it could be incomplete - Doesn't account for crime funnel

Answer 70

A statistically based literature review.

Answer 71

A type of review article, in which findings from individual studies are pooled with other findings from similar studies. The studies should always be accompanied by a systematic review.

Answer 72

Of crimes committed, not all are reported, of which not all are investigated, of which not all are prosecuted, of which not all are convicted- which affects data.

Answer 73

Archival data are often aggregated- they don't look at individuals, rather groups, making inferences about individuals based on results from groups. This is an incorrect assumption- relationships that apply at group level do not always apply on an individual level.

Answer 74

A method of summarising and synthesising primary studies on a topic, without a premeditated research question, often contributed to by the researcher's own experience and existing bias.

Answer 75

A method of summarising and synthesising primary studies on a topic. Authors have been careful to be more exhaustive than in narrative reviews. They are different from narrative reviews, because they: - Have a protocol for literature search and data extraction - Aim to be comprehensive - Appraise the quality of a study

Answer 76

The standard of differentiation between two means. Usually represented in forest plots.

Answer 77

Putting together findings of different data sources and analysing them to gain understanding. (Different to meta and review analyses, which are more "book" orientated).

Answer 78

A data collection method. The investigator simply observes-

Answer 79

1. Cohort studies 2. Cross-sectional studies 3. Case-control studies

Answer 80

A type of observation study. They describe incidence or natural history. They measure events in chronological order, so they can be used to find causes and effects. They may be prospective or retrospective.

Answer 81

A type of cohort study, which is a type of observation study. People in the sample are observed to see whether they develop the outcome of interest.

Answer 82

Pros: - Can measure causes separately from effects, so there is no debate - A single study can examine various outcome variables - Can calculate the effect of variables on the probability of developing the outcome Cons: - Subjects don't always follow up - The rarer the condition studied, the harder it can be to determine causes and effects.

Answer 83

A type of cohort study, which is a type of observational study. They use data already collected for other purposes. People in the sample are observed to see whether they developed the outcome of interest. The cohort is followed up retrospectively.

Answer 84

Pros: - Cheaper, since data is already collected - Lack of bias, since study was for different purposes Cons: - Unlikely that all relevant information will have been collected - Following up with people later can lead to recall bias

Answer 85

When two cohorts are compared, one will be exposed to the agent of interest and one will not. However, it is not always possible to control all other factors that differ between groups. This means that confounding variables occur, which are associated with both the exposure and outcome that is not the studied variable. Confounding variables are a threat to the internal validity of a study, or the extent to which its results indicate a cause-and-effect relationship between X and Y.

Answer 86

A type of observation study. Used to determine prevalence. They do not permit distinctions between cause and effect. All measurements on each person are made at one point in time. They can be used to infer causation. Subjects are not exposed/treated, so there are seldom ethical difficulties.

Answer 87

Pros: - Quick and cheap - Best way to determine prevalence - Useful at identifying associations Cons: - Cannot differentiate cause and effect - Often multiple plausible explanations - Rare conditions cannot be effectively studied

Answer 88

A type of observation study. They compare groups retrospectively. They seek to identify possible predictors of an outcome. Used to generate hypotheses that can be studied via prospective cohort or other studies. People with the outcome of interest are matched to a control group without it. Retrospectively, the researcher determines which individuals were exposed to the agent.

Answer 89

Pros: - Good for rare conditions - Only option when there is a long time between exposure and outcome - Fewer subjects required to test - Useful for generating hypotheses for other studies Cons: - Restricted outcomes- the only outcomes possible are the presence or absence of the subject of study - Big concerns surrounding bias- sampling bias and observation and recall bias

Answer 90

The bias within test or control groups. Sampling bias occurs when a sample used in a study or experiment is not representative of the population from which it was drawn.

Answer 91

Biased assessment of (predictor) variables, their presence and significance, by the subject and/or the investigator.

Answer 92

A period of time between exposure to an agent and the "development of symptoms", signs or other evidence of changes associated with exposure.

Answer 93

The ratio of the probability of an event occurring to the probability of the non-occurrence.

Answer 94

The ratio of the probability of developing the condition if exposed to a certain variable, compared with the probability if not exposed.

Answer 95

The rigour with which a study has been designed and executed- that is, can the experiment be relied upon? It largely depends on study design.

Answer 96

The usefulness of findings of a study with respect to other populations.

Answer 97

A variable that is thought to cause the dependent variable.

Answer 98

A variable that is thought to be caused by the independent variable.

Answer 99

1. X must occur before Y (temporarily) 2. X is statistically related to Y (there has to be statistical evidence that they are related) 3. No confounding variables involved (the statistical relationship that we find between X and Y is not because there is a causal relationship between X and Y, but because of factor Z)

Answer 100

A research method where the researcher manipulates one or more independent variables to observe their effect on the dependent variable, usually within controlled conditions. This design helps establish cause-and-effect relationships by isolating the variable of interest and minimising the influence of confounding variables. Participants are often randomly assigned to groups.

Answer 101

1. The experimental group 2. The control group

Answer 102

A group within an experimental design. Comprised of participants who receive a certain intervention. The difference between the results of the experimental group and the control group will show how effective the intervention was. People are often blind to the knowledge of which group they are in, to avoid bias.

Answer 103

A group within an experimental design. Comprised of participants who do not receive an intervention, or receive a different one. The difference between the results of the experimental group and the control group will show how effective the intervention was. People are often blind to the knowledge of which group they are in, to avoid bias.

Answer 104

Random determination of who is in the control group v who is in the experimental group. If, after random assignment, you can still see similarities amongst people within certain groups, ignore it- random assignment is trustworthy enough to not need to change the groups further.

Answer 105

1. The independent variable is manipulated by the researcher 2. Experimental and control group involved 3. Random assignment of participants 4. Measurement of the dependent variable before and after the independent variable is introduced.

Answer 106

In criminology, the classic experimental design often cannot be used because of practical or ethical constraints. An alternative design, quasi-experimental, is possible. (Though most criminological research is observational). In quasi-experimental designs, the independent variable is still controlled by the researchers, but one or more of the features of the classical experimental design is not present. (i.e. there may be no control group (reversal design), random assignment or pretest (e.g. after-only design).

Answer 107

An example of a quasi-experimental design. 1. Independent variable is manipulated by the researcher 2. Only an experimental group (experimental and control conditions occur within same participants) 3. No random assignment (same participants undergo all phases) 4. The dependent variable is repeatedly measured under experimental and control conditions, before, during and after the independent variable is introduced (rather than having 2 control groups, there is one that sometimes gets it and sometimes doesn't)

Answer 108

An example of a quasi-experimental design. This is a design which skips the pretest, and only measures the dependent variable after the intervention. 1. Independent variable is manipulated by the researcher 2. Experimental and control groups are involved 3. Random assignment of participants 4. Measurement of the dependent variable is taken only after the independent variable is introduced (no pretest)

Answer 109

- Withholding/not giving every person the item or treatment can be bad, e.g. withholding medication - There is no account taken of the learning effect (i.e. that people will learn from the experiment and change accordingly during it)

Answer 110

- When you don't measure beforehand, you can't be sure of a lack of bias - Placebo effect - Retesting bias - Limiting ecological validity (application in the real world) - Hawthorne effect - Spill-over effect

Answer 111

If you test people before, and then test them after, they may remember how they previously responded, which can influence results

Answer 112

Knowing that you are participating in a study and being watched can influence your behaviour

Answer 113

When a particular intervention is introduced to a group, they may still interact with other experimental or control groups, meaning that your intervention may be received in those groups, too.

Answer 114

Refers to a type of research method where the researcher observes and records behaviour without manipulating any variables.

Answer 115

Top: Reviews - Meta-analysis - Systematic review - Narrative review Middle: Experiment - Classical - Quasi Bottom: Observational - Cohort - Case-control - Cross-sectional

Answer 116

A group of people who share one or more characteristics that are of interest to the researcher

Answer 117

A subset of a population

Answer 118

A sample whose members possess all characteristics in the same proportion as the population

Answer 119

A sampling method, in which: - Each person has the same chance of being selected - Sample can be considered representative - Sampling frame required

Answer 120

A list of everyone in the population

Answer 121

A sampling method, in which: - Some people have a higher chance of being selected than others - Sample cannot a priori be considered representative (unless checked after the fact) - No sampling frame - Generally seen as weaker than probability sampling methods- you shouldn't non-probability methods when you're trying to generalise your findings to a wider population

Answer 122

Refers to knowledge or justification that is independent of experience. Truths can be known through reasoning rather than empirical observation.

Answer 123

1. Simple random sample 2. Stratified sample 3. Systematic sample 4. Cluster sample

Answer 124

1. Convenience sample 2. Quota sample 3. Snowball sample

Answer 125

Randomly picking from the population, and everyone that you randomly pick belongs to the sample. Everyone has an equal chance of being included in the sample (but this is difficult to do).

Answer 126

You have a population, within which some subgroups are not the same size (e.g. more men than women). In this case, you split your population into strata (subpopulations), and then randomly select from each strata- this allows for everyone to be included.

Answer 127

Within stratified samples, you have disproportional stratified samples, which means including an equal amount of people in your sample- this means there is no longer an equal chance of being included (more women would be included than men based on how many were originally available).

Answer 128

You still have strata, but there is a proportionate amount of each subgroup that was in the population. Better estimates of the actual population, but if you want to know more about the smaller subgroup, you are at a disadvantage.

Answer 129

Subpopulations within a population

Answer 130

Sampling frame is laid out in a row, and every kth entity is sampled. This means that the individual chances of being selected are not the same, since if you are not divisible by k, you stand no chance. If you have 100 people and you need 3, you should not take the 3rd, 6th, and 9th. You should split it equally throughout the population, so the ones at the end are also represented. Requirements: - Random place to start - No periodicity with the same frequency as k (no patterns in the data that are reflected in the same pattern as k- so every k happens to also be German/pregnant/male/under aged 50 etc.)

Answer 131

Sampling of naturally occurring groups within a population (e.g. prisons, schools, classes, sports clubs, cities etc.) From these multiple clusters, you select a random sample- then everyone within that final cluster will be included in your sample.

Answer 132

Individuals are selected because they can be easily reached. An example would be that a lot of psychology studies are done using first-year psychology students. As long as you report on what your sample looks like when sharing your research results, this can be ok.

Answer 133

Non-probabilistic equivalent of stratified sampling. You create subpopulations and draw samples, but they are less purposely random. You draw a non-probability sample from each subpopulation. A cousin of stratified, just that stratified is based on probability and quota is not.

Answer 134

Respondents help to recruit new respondents. Used when respondents are difficult to locate, or unlikely/unwilling to participate. This could make this less random- since they know each other, and are likely to have commonalities.

Answer 135

- Representativeness (it will be a better representation of the wider population) - Statistical power (making sure you have enough data to find a difference where there is a difference/to be able to conclude something meaningful) - Precision (the more precise your estimates will be)

Answer 136

Those who are selected and invited to participate in the study

Answer 137

Participants who participate in a study (not all who are chosen participate)

Answer 138

The difference between the chosen sample and the obtained sample.

Answer 139

- Willingness to participate - Inability to participate - Participants cannot be contacted - Participants do not follow the study protocol

Answer 140

Participants who do not participate (anymore) differ systematically from those who do. This can be a threat to the internal validity (are our results measuring what we say they are measuring? You may not have an accurate/relevant study group to successfully measure your question) and external validity (how much it can be attributed to the wider population) of a study. If you see that this occurs, you can still publish results- as long as you qualify your results by saying they only apply to the cohort that actually participated in the study.

Answer 141

- Notifying respondents beforehand - Including a polite and clear introduction - Personalising the correspondence - Explaining the importance of your study - Having a well-respected or well-liked sponsor - Keep the questionnaire short - Making sure the questionnaire is neatly presented - Compensating participants (this may affect the type of people that take part, though)

Answer 142

The process of making an abstract concept (the construct) measurable (as a variable) for a study.

Answer 143

1. Identifying the construct 2. Defining the construct (conceptual definition) 3. Conceiving a method to measure the conceptual definition (operational definition)

Answer 144

1. Nominal 2. Ordinal 3. Interval 4. Ratio

Answer 145

Categories that cannot be ranked. E.g. countries, gender, eye colour, type of crime

Answer 146

Categories have a logical order but no quantifiable distances between them. E.g. opinions (strongly agree, agree etc.), Olympic medals, severity of crimes

Answer 147

Categories have a logical order and have equal distances between them. E.g. acidity, temperature

Answer 148

Categories have a logical order and have equal distances between them, and there is a meaningful 0 (i.e. nothing). E.g. income, age

Answer 149

The overall consistency of a measure. Based on the idea that our observations have a true value + some kind of measurement error.

Answer 150

The degree to which one is measuring what is supposed to be measured (e.g. the validity of an IQ test giving valid indications on the level of intelligence)

Answer 151

1. Measurement error caused by differences between assessors, inter-rater reliability 2. Measurement error caused by situational awareness, test-retest reliability 3. Measurement error caused by the instrument itself, split-half reliability

Answer 152

Agreement between measurements taken with the same instrument by multiple assessors. Then cross referencing results between assessors into a table (see slide) to see where conclusions are the same/differ. Typically, the reliability is then calculated.

Answer 153

Agreement between measurements taken with the same instrument on multiple occasions. Certain conditions should be the same, including observers, location, time of day. Would hope that when taking a test again, the result would remain the same.

Answer 154

Agreement between measurements taken with the two halves of an instrument, (e.g. even- vs odd-numbered items). A measure of internal consistency, or the extent to which the items of an instrument measure one construct.

Answer 155

The suitability of an instrument based on overall appearance. If you want to measure the criminal tendency of people, weighing them would be inappropriate- but we need to keep our minds open to more creative ways of measuring.

Answer 156

The extent to which an instrument covers all aspects of the construct.

Answer 157

The correspondence between scores on instruments designed to measure the same construct.

Answer 158

- Assurance of anonymity or confidentiality - Clearly phrased questions - Give examples - Non-judgemental phrasing or prefacing of questions - Short reference periods - ‘Warm-up’ questions

Answer 159

- Academic: journals, books, conferences - Non-academic: policy briefs, government reports, social media posts, blogs, public lectures, media interviews, stakeholder meetings

Answer 160

1. Research tries to increase our understanding of something (the problem) 2. Through this, the research seeks to improve our lives (the solution) 3. Effective communication is vital to ensure that research findings reach the people who can implement them (knowledge transfer)

Answer 161

People who have an interest in your research. Potential stakeholders include academics, policy makers, professionals, target group, general public

Answer 162

1. Transparency 2. Replication

Answer 163

1. Introduction - explains the "problem" that the study tries to resolve 2. Methods - describes the actions (e.g. data collection, study design, analyses) taken to resolve the "problem" 3. Results - presents the findings from the analyses, which forms the basis of the "solution" to the "problem" 4. Discussion - interprets the findings and makes recommendations for practice (the "solution") and future research Other components: - References - Conflict of interest - Author contributions - Acknowledgements - Title (and subtitle) - Authors and affiliations - Abstract

Answer 164

- In text - Reference list

Answer 165

- Involvement of people or animals - Environmental consequences - Impacts on society or future generations

Answer 166

- International: Declaration of Helsinki - National: Codes of ethics - Local: ethics committees

Answer 167

- Informed consent - No deception - Minimising harm - Anonymity or confidentiality - Approved by ethics committee

Answer 168

A principle of ethical research. Prospective subjects must be fully informed about the study before they agree to participate. This can be: - Explicit: Signed by the subject - Implicit: Anonymous questionnaires

Answer 169

A principle of ethical research. Deception cannot be allowed, either: - by omission: giving incomplete information - "orchestrated": instructions that cause subjects to believe something that is not true (there are exceptions)

Answer 170

- Observations in public spaces (can be difficult to contact people in public etc.) - Ethnographic studies (observations of people’s behaviour, “infiltrating” communities) - Experiments, under certain (strict) conditions (give some information but not the whole truth, for the purpose of not ruining the study. Fully debrief afterwards!)

Answer 171

The researcher does not know who the participants are

Answer 172

The researcher knows who the participants are, but removes identifying information from the dataset and report

Answer 173

- Train all researchers in research ethics - Restrict access to the participants and their data - Remove identifying information from data files - Store files securely (e.g. locked cabinets, encryption)

Answer 174

Oversee research involving human participants. They: - Are local and field-specific - Review proposals for research - Are accredited, but generally supervised

Answer 175

Typically has to do with reporting your research. “Intention or gross negligence leading to fabrication of the scientific message, or a false credit or emphasis given to a scientist”

Answer 176

- Plagiarism - Fabrication of data - Falsification of data

Answer 177

Intentionally presenting someone else's writing or ideas as your own.

Answer 178

1. Copying passages without credit 2. Ghost-writing 3. Improper authorship

Answer 179

Making up data. Includes the intentional use of references that are fake or do not support an argument.

Answer 180

Misrepresenting data. Manipulating results- for example, you collected real data, but you leave certain people out of your sample so that it answers a question in the way that you want.