Ethics and Methods in Psychology PDF

Ethics and methods Introduction: methodological and ethical issues in the Milgram experiment  The aftermath of the Milgram experiment The experiment had certain flaws...  Lab. exp.: low external validity, non- generalizable.  Biased sample: only male participants in the orig...

Ethics and methods Introduction: methodological and ethical issues in the Milgram experiment  The aftermath of the Milgram experiment The experiment had certain flaws...  Lab. exp.: low external validity, non- generalizable.  Biased sample: only male participants in the original experiment.  Participants might have quit; the experimenter pushed the „teachers”. But it still changed how we think about obedience and responsibility, and Contributed to the development of the ethical dimension of research in social psychology.  Ethical issues Causing unacceptable level of stress to participants: Teachers: nervous laughter, nails in the arm, fit of giggles Harms the self-evaluation of participants Long-term psychological effects: subjects perceive that they are capable of killing another human being in specific situations. Participants might feel being abused and being ridiculous when the experimenter eventually informed them about the real aim of the experiment. Abusing the prestige of science Risking the recpect for psychologists in particular and scientists in general. Experiments in social psychology and ethics Freedom of scientific inquiry ↔ people’s right to human dignity and privacy Questions: Are the social benefits of such an experiment higher than its social costs? Main ethical concerns Deception  Basicly a lie  Which experiment needs it and why?  Does it justify deception? (fundamental ethical dilemma) Harmful effects  Physical harm: e.g., pain, inconvinience  Psychological harm: e.g., harm of dignity, boredom, distress, self-knowledge(!) Ethical codes and committees  E.g.,: Ethical code of the Hungarian Psychological Association (HPA)  Ethical committees at different institutions (e.g.,. Ethical Committee of the HPA)  The ethical committee has the right to refuse the research plan or to ask for amendments Ethical principles  Principle of protection: in connection to the experiment, the researcher must protect participants from physical and mental harm higher than what they experience in ordinary life.  Principle of informed consent: the researcher should explain in advance what is involved in the experiment (so far as is practicable), and should obtain the informed consent of participants.  Principle of confidentiality: the researcher must respect the participants right to privacy; data must be kept anonymous unless the participants give their full consent.  When this requirement cannot be met?  When informing participants about the real aim or the hypotheses of the experiment would influence the output – deception.  Giving aid to others  Ethnic or gender prejudice  Aggression  Conformity  Debriefing A protocol/discussion after the experiment when the experimenter gives participants full and detailed information about the experiment including the planning phase, the aims and the results. If the experimenter used deception she should reveal the ‘real deal’ behind the cover story, and, should explain why it was necessary to use deception – by doing this she also checks whether the deception worked. Aims:  To ensure that participants leave healthy and with good sentiments  To foster participants to learn something about themselves, psychology and humans in general.  To foster experimenters to learn  To ensure that the cover story worked Does regulation impede research? Milgram reloaded: Jerry Burger conducted a remake of the Milgram experiment (2009) with a twist: he followed all contemporary ethical principles  Preliminary psychological analysis of applicants: ensured that all participants have full mental health  Reminded the participants multiple times that they can quit the experiment without losing the 50$ fee.  Debriefing starts with the truth (Milgram revealed it later)  The experiment ends with 150 V 70% of the participants obeyed (basicly the same as Milgram’s (82%) Male and female participants Social psychologist vs. layperson Social psychologist vs. layperson: cognition The difference is not in the aims but in the ways! Different approach. e.g., in judging the happiness of those winning the lottery or get promotion, and the unhappiness of those having serious illness. Conclusions based on scientific research do not always fit to our common sense We are biased without knowing it, and without being able to defend ourselves In later lessons: schemes of information processing, heuristics Everyday and scientific cognition Everyday Scientific Imprecise, „undisciplined” Eliminating perception biases perception Critical stance Overgeneralizations, there Testing theories with is no systematic inquiry, empirical data perfect logical structure is Rational, logical and coherent not an aim conclusions Non-rational conclusions Automatic processing Systematic processing Heuristics Aim: Rational, objective Aim: Practicability statements What is social psychology? In a nutshell: science of social behavior In a larger nutshell: a science which explores how the thoughts, feelings and behavior of individuals are influenced by others (Allport, 1985) What is social psychology good for? It aims to make observations – preferably based on sense perception – and to make, from such observations, hypotheses to be tested. Scientific explanations about the past and present of the social world, and forecasts about its future. Not a perfect, but an increasingly sophisticated tool to develop awareness in our social relations and to influence them to the desired direction. What is social psychology not good for? Does not provide ‘once-and-for-all’ descriptions of the world. Cannot be used as a handbook (not a ‘life hack for dummies’). It is not possible or desirable to reduce the complexity of our social world to a couple of scientific theories. It is not a cumulation of folk wisdom or the lessons learnt through social life but a scientific discipline with all its social characteristics. More knowledge of social psychology does not make our life easier or better in itself, but it can make it better if properly used. Methods Methodology of empirical sciences Empirical science: primarily aims to draw conclusions from (direct) sense experience. Builds theories on perception and data; seeks co-occurrences, correlations and cause-and- effect relations in social behavior. Aims to provide explanations and forecasts about the psychology behind the behavior of human communities. Makes hypotheses based on observations, tries to do systematic testing (justify or falsify)  Sources of data Social psychology – like any other science – collects data through making observations in a systematic way: Main methods of collecting data: laboratory experiments, field experiments, correlation studies, descriptive studies, participant observations, tests, interviews, surveys, longitudinal studies...  Experimenting 101 Experimenting requires to identify the variables (i), and to explore the causal relations between them (ii):  Independent variables are those which we manipulate in a systematic way to induce an observable change in the values of dependent variables (independent variable ~ cause); manipulation.  Dependent variables are those in which we tries to induce an observable change by manipulating the independent variables (dependent variable~ effect); measure.  Extraneous variables might be a source of error therefore their values have to be fixed in different variations of the experiment; control. Ceteris paribus: with other conditions remaining the same. All variables remain fixed but one independent variable which we manipulate – we observe how this change influence the value of dependent variable(s). On experimental social psychology Random allocation: randomly allocating participants of the experiment to independent variable conditions. Participants have equal chance of taking part in each condition. Aim: limiting the effect of participant variables; providing higher control. Internal validity: the degree to which the experiment provides causal information between independent and dependent variables (higher in experimental research, lower in correlational research). External validity: the degree to which results of the experiment might be generalized beyond the experimental sample, setting and situation. Laboratory experiment This experiment is conducted in a well-controlled environment, independent variables are manipulated by the researcher (e.g., Asch and Milgram experiment) Advantage  Relatively easy to control independent and extraneous variables  Easy reproducability Disadvantage  The artificial manner of the experimental setting might affect the results of the experiment  Participants might be affected by different social influences than in natural settings. With other words, disadvantages are: low external validity and high influence of demand characteristics.  Field experiment This experiment is conducted in the everyday environment of participants, independent variables are manipulated by the researcher (Muzafer and Carolyn Sherif ’s experiment with children) Advantage  Behavior of participants is more likely to follow real-life patterns; low influence of demand characteristics because participants often do not even know that they are participating in an experiment and/or being observed.  High external validity Disadvantage  Harder to control extraneous variables  Less appropriate for manipulating one independent variable and monitoring its effects (better for groups of variables)  It might last longer because the researcher cannot „make the scene” but has to wait for the appropriate circumstances to naturally occur.  Natural experiment This experiment is conducted in a natural environment, independent variables are not manipulated by the researcher Advantage  The behavior of participants is the most likely to follow real-life patterns in this type of experiment.  Low influence of demand characteristics, participants often do not know that they are participating in an experiment.  Might be used in cases when manipulating independent variables by a researcher might raise serious ethical concerns. Disadvantage  Very expensive and long  No control over extraneous variables – impossible to reproduce the experiment  Low internal and external validity, low influence of demand characteristics Correlation  On correlation studies Correlation method can be used to analyze the co- occurrence of two variables. To be used to analyze the relation between two numerical variables The sign of the correlational coefficient (r) shows the direction of the relation, the magnitude of the relation (between 0 and 1) shows the consistency of their co-occurrence, the power of the relation.  If the correlational coefficient is 0, there is no correlation  If the correlational coefficient is 1, there is perfect correlation  If the correlational coefficient is -1, there is also perfect correlation, but the direction of the relation is inverse.  Correlation and causality  But correlation does not necessarily mean causal relation between variables  e.g., there is correlation between the softness of asphalt and the number of apopleptic strokes  But its not the fumes from the soft and warm asphalt which poison us causing stroke – there is a third variable, the hot and shiny summer weather which causes both.  High correlation between the number of storks appearing and nesting in French villages and the number of human birth-givings – is there a causal relation?  The more variables we take into account, the more probable it is that we find some statistical relation between them.  Like the length of names and years of education or the number of „t”-s in test questions and the letter of the good answer (t – truth, in a truth/false)A– but these are coincidental correlations Be careful with correlation! Correlation is not a causal relationship! Types of test Multiple choice test Statement (e.g., I never do anything dangerous, I rarely dream daytime, sometimes I think faster than I can express my thoughts); true, not typical, I do not know, or abcd (e.g., IQ) Projective test e.g., Rorschach (although it is not a test of social psychology but a clinical test; 10 figures, partly coloured, symmetric, „butterfly”-like ink blobs)  There is no context being given, undefined stimulus, plethora of possible responses, but there are typical answers (the Rorschach test should be conducted by a qualified expert!)  Half-projective test  There is some context, but the stimulus is underdetermined. e.g., interpretation of drawings, complete sentences, arranging photos, etc.  Data vs. theories  Finite data can always be explained by multiple theories – general trait of empirical sciences.  Usually the debate is not about the phenomenon (data) but how to explain it: e.g., Zajon’s motivational drive theory and Baron’s attention theory explain the same experiments.  According to Zajonc, the presence of others cause higher drive/arousal which calls forth automatically the dominant answer – inadequate, in case of complex tasks. Baron thought that, although the presence of others raise our motivational level it also distracts part of our attention. The more simple a task is, the less attention is needed thus we are doing better in simple tasks and worse in complex tasks in the presence of others.. Demonstration Thinking Critically About Causality Give at least one alternative interpretation of the articles (British scientists) on the handout! Qualitative methods Qualitative research A protocol which fosters the contextualized, uniqueness-driven description and understanding of the phenomena. Inductive research strategy Higher emphasis on interpretation Analysis of real-life events Non-representative samples More attention to extremities Methods  observation  content analysis  biographical analysis  qualitative interview  focus group  case study  Qualitative research: Pros and Cons Advantage Enables to collect rich and essential information. Be able to grasp unique situations in a complex way Useful in the initial (explorative) or in the ending phase (case study) of research Examine phenomena in their context Disadvantage Their reliability and validity is doubtful (cannot be reproduced, rarely anecdotic, not objective) Observation: Pros and Cons Technique of recording: video or voice recording, reports, observational journal, writing research log When to use? In the early phase (pilot analysis) If there is no other way of making inquiries In case of sensitive topics Advantage Be conducted in natural settings, good ecological validity Less or no intervention Disadvantage Time- and labour-consuming; experience is crucial Risk of subjectivity A threat to reliability: the observation might affect what is being observed Cannot be used to explore cause and effect relations  Content analysis Content analysis is a research technique which draw conclusions from a text by analysing its traits in a systematic and objective way (Stone, 1966) Content analyis can be used on paintings, newspapers, Tv shows, speeches, acts, websites, basicly anything... Related concepts: discourse and speech act analysis, document analysis, comparative historical analysis, analysis of available statistical data Content analysis: Pros and Cons Advantages Disadvantages Capable of handling large Difficulties in making and using amount of unstructured data distinct categories; assuring the Capable of handling data of a reliance of categorization long period Time-consuming Non-interventionist Can only be used on previously technique recorded communication Sensitive to context Tendency for oversimplification Adequate for qualitative The technique is often limited to analysis simply counting words Capable of analysing events in a large time span or a broad geographic spread Interview Depth interview (personality) Narrative interview (biography) etnographic (culture) etnomethodology (constructing reality) thematic (given theme) Advantages complex, diverse information flexible, „tailor-made” Disadvantages Time- and labour-consuming Low generalisability Case study Thorough study of one or more instance of a social phenomenon. Comprehensive research strategy which consists multiple research methods (interview, test …) Processing one or more cases The analysis can be done on individual or on social level Question of authenticity  On the level of data  On the level of interpretation  Focus group Group interview (6–12 people)  A given theme  The role of moderator (facilitating interaction, controlling the debate, making positive atmosphere)  participants – selection is driven by the theme → question of homogeneity  Used mostly in marketing research  Rapidly developing area: online focus group.  Functions and significance Functions Exploring the problem Making hypotheses Developing questionnaires - validation Evaluating the program, impact assessment Interview vs. asking the individual in a group Group dynamics (agreement – difference in opinion, dynamics of minority – majority) Shared and commonly built interpretations (the responses are not simply complementary but developing a dynamic relation with each other) Emotional, non-conscious responses might appear Evaluating the method Advantages flexibility fast, simple to use Disadvantages It is difficult to analyse data Question of generalisability Selection of group members  Summary  Social psycology is an empirical science which collects data in a systematic way by using experiments, questionnaires and other research methods.  Theories are based on discovering cause-effect relations. When testing hypotheses, the aim is to identify and to monitor independent and dependent variables as precisely as possible. (various methods of collecting data are differently useful in this respect)  To differentiate between cause and effect in case of human behavior, it is best to use control groups. When testing a hypothesis, members of the control group are influenced by every variable except the independent variable. (Members of the experimental group are influenced by every variable.)  In certain cases, it is hard to tell a cause-effect relation from a correlation. Correlation is the co-occurrence of two variables but it is not necessarily a symptom of causal relation. Literature Smith, Eliot R. - Mackie, Diana M. – Claypool, Heather M. (2014) Social Psychology (4th edition), New York: Psychology Press. Babbie, E. (2018) The practice of Social Research, New York: Wadsworth. Szokolszky, Ágnes (2004) Kutatómunka a pszichológiában. [Research in Psychology] Budapest: Osiris Kiadó.