JoHo & Leren
- Waar kan ik JoHo vinden?
- Waarom & hoe donateur worden?
- Waarom & hoe serviceabonnee worden?
- Waaraan denken voor en na je eindexamen?
- Waar & hoe stagelopen in Nederland?
- Waar en hoe studeren in het buitenland?
- Waar en hoe stagelopen in het buitenland?
- Waar en hoe een taalcursus- of reis doen?
- Waar en hoe vind ik een bijbaan of tijdelijk werk?
- Waar kan ik meer vinden over de TEFL cursus?
- UVA Psychologie Amsterdam
- VU Psychologie Amsterdam
RUG Psychologie Groningen
UL Psychologie Leiden
- RU Psychologie Nijmegen
- EUR Psychologie Rotterdam
- UU Psychologie Utrecht
- TiU Psychologie Tilburg
Pedagogiek & OK
Bedrijfskunde en International Business
- Business Economics
- Sociale wetenschappen
- Statistiek & Methoden en Technieken
- Samenvattingen gebruiken na je studie
- Studiehulp zoeken
- Waar en hoe vind ik een stage in het buitenland?
- Waar vind ik stages per werkveld?
- Waar vind ik stages die bij mijn opleiding passen?
- Welke organisaties kan ik het beste inschakelen of laten bemiddelen?
- Waar moet ik opletten als ik stage ga lopen in het buitenland?
- Stagelopen heeft invloed op mijn verzekeringssituatie .. hoe zit dat?
- Waar vind ik extra fondsen of financiering?
- Voor abonnees
- Voor donateurs
- Bijbanen & Studentenbanen
- Sabbatical nemen en carrière switch zoeken
- Hoe maak ik gebruik van de vacatureservice?
- Hoe kan je je competenties & kwaliteiten checken?
- Hoe maak je werken leuk?
- Hoe vind ik werk dat bij mij past?
- Wat leer ik van tijdelijk werk in het buitenland?
- Samenvattingen zoeken voor je oriëntatie
- Voor donateurs
Summary An introduction to Social Psychology (Hewstone)
Te gebruiken bij
Auteur(s): Hewstone, M., Stroebe, W.
Druk/Jaar van uitgave: 6/2015
Zoek recente samenvattingen & studiehulp
2. Research methods
Facts, Theories, Hypothesis
A fact is an objective statement usually based on direct observation that can reasonably be accepted as true. A theory is a conceptual model (idea) that attempts to explain known facts and make predictions about new potential facts. These predictions are called hypotheses.
Psychology differs from sociology, philosophy and anthropology in its method of research (the scientific method). Psychology tests the theories extensively, while the other disciplines don’t do that in the same way. Theories need to be testable. A theory can get stronger with evidence, but it can also get weaker or even scrapped by counter-evidence. Also, theories can get boundary conditions. This means that the theory only predicts something under certain conditions.
Three Important Lessons
Scepticism towards both miraculous claims and reasonable-sounding scientific theories maintains scientific integrity.
Careful observations must be made in controlled conditions for results to be taken as having a reasonable reliability.
Observer-expectancy effects are those in which the subtle behaviour of the observer indicates how the subject acts.
There are three main categories of research strategies:
Research design, which includes experiments, correlational studies, and descriptive studies.
Setting in which the study is conducted, either in the field or in the laboratory.
Data-collection, through self-report and observation.
Experiments are used to test a hypothesis about a cause-effect relationship between two variables. The variable that is the supposed cause is the independent variable, and the supposed effect occurs to the dependent variable. The independent variable can be manipulated to observe how the effect on the dependent variable differs. When all other variables are constant, it is easiest to observe real effects on the dependent variable and establish causation. People and animals that are studied in a research are called subjects. There are two types within an experiment, namely the quasi-experiment and the true randomized experiment. In the quasi-experiment, participants aren’t distributed randomly to the different experimental conditions. That’s because quasi-experiments take place in the natural setting and the researcher can’t really control the natural setting. A true randomized experiment is an experiment in which participants are distributed randomly to the different experimental conditions. The researcher has complete control over the independent variable(s) and the participant. True randomized experiments are often done in the laboratory, so the researcher can control everything.
In within-subject experiments, each subject is studied within different conditions of the independent variable. That is, many people may participate in the experiment, each tested with the different conditions of the independent variable.
In between-groups experiments, there is a separate group of subjects for each different condition of the independent variable. This usually also includes a control group, for whom the conditions are normal – this allows for data that can be compared against. In between-group experiments, random assignment is used to determine which group subjects will belong to. This helps control against possible confounding variables like age, sex, birthplace, etc. as well as the way people are treated.
In some cases, ethical and practical reasons prevent the conducting of experiments. A correlational study is one in which the researcher does not manipulate any variable, but instead measures two or more existing variables to determine relationships between them.
Correlation versus Causation
While it is tempting to treat correlational results as if one variable caused the other, without controlling the variables, it is not possible to determine causation. Causal relationships may go in two directions, or the reverse of what is assumed. There may also be a third, unknown variable which lies at the heart of the observed correlation.
These studies aim to describe the behaviour of an individual or set of individuals, without assessing relationships between variables. These studies may or may not involve numbers. They can be narrow in focus, looking at one aspect of behaviour, or broader, aiming to learn as much about one group or individual as possible.
Conducting research in a laboratory allows data to be collected under controlled conditions. However, the clinical and abnormal atmosphere of a laboratory may have an effect on the subject’s behaviour. The results may not reflect reality.
Any research conducted outside of the laboratory is called field research. These settings may include the subjects’ workplaces, homes, consumer areas, or other parts of the subjects’ normal environments. This has the disadvantage of being difficult/impossible to control, but the advantage of providing more reality-based results.
A survey is a research strategy that makes use of interviews and questionnaires. There is no manipulation in this: existing variables are measured. Before conclusion can be drawn, other variables also need to be tested. Otherwise you won’t know if the found relationship between the independent and dependent variable is really there.
Procedures in which people are asked directly to reflect and report on their own mental states, attitudes and behaviour, often done through a written questionnaire or an oral interview. Questionnaires are a fast, cheap and easy method. Also, many behaviours can’t be observed, so this might give researchers a bigger change to look into certain non-observable behaviours. During an interview, the participant can be recorded (if he allows this). This give the researcher the chance to listen back to the answers. A disadvantage of an interview is that it’s quite expensive and costs quite a bit. A disadvantage of the questionnaire is that only 10 to 50% of the participants fills the questionnaire in and sends it back. Also, people can always lie during interviews and in questionnaires.
Observational procedures are those by which researchers observe and record behaviour without self-report. This type of data collection includes naturalistic observation, when the researcher avoids interacting with the subjects. It also includes tests, in which the researcher presents problems or situations to which the subject must respond. When people know they are being observed, they may act differently and this is called reactivity.
An internet experiment is an experiment that is conducted on the internet. In other words, with a computer. An advantage of this is that you can gather a great amount of data in a short period of time. It’s quite easy. Also, you can reach people from different countries, which is good, because people from different countries often have other cultures and you can see whether your study results are generalizable. A disadvantage is that the researcher doesn’t have control over the study, which results in participants doing the study under difference circumstances. Another disadvantage is that results can be flawed, because the study may be conducted in another language than the native language of the participants, which may result in them not fully understanding and comprehending the question. Also, the participants are volunteers, which may not be so representative for the entire population. Some researchers think these things are not really big problems and that internet experiments are not that different from normal experiments. Internet experiments will therefore be used more and more.
Implicit measures measure unconscious constructs, like reaction time. These measures can tell us more about cognitive processes, which individuals are often not conscious about. These implicit measures are often done with computers, which gives accuracy.
In social neuroscience, the neural activity is measured. This approach assumes that all human behavior is biologically determined. The neural activity is measured with fMRI scans, which measure the blood-flow related to neural activity.
A one-shot case study looks like an experiment, but isn’t like one. You really observe a certain phenomenon extensively, but you can’t compare it to something. One example is when a child is found, living with wolves and you decide to study that child. You can’t really compare that child to another child that has been reared by wolves, because there aren’t other children reared by wolves. So the conclusion you draw, may actually be flawed.
It’s actually important to use control groups. When you test a medicine, you may give one group who has diabetes the new medication, while you don’t give medication to the other group (the control group). If the first group indeed gets better, it’s safer to say that the medication works. If it really turns out that the medication works, you also have to give it to the control group after the experiment has been conducted, otherwise it’s unfair for them.
However, sometimes it seems that there’s a relationship between two variables and that the one causes the other, but the relationship turns out to be actually caused by a third variable. A moderator effect occurs when there is only an effect when a third variable is present. If the third variable isn’t present, than the first variable doesn’t have an effect on the second variable. Mediation means there is an indirect effect. This means that the first variable has an influence on the third variable, which in turn has an influence on the second variable. So, the first variable influences the second variable indirectly.
Research can be qualitative or quantitative. Qualitative means that it looks at words, provided by the participant. The content of the words is very important and because of this, interpretations can vary. Quantitative research looks at data scores, which can be interpreted. Most studies are quantitative. According to some scientists, it’s best to perform multiple research methods, which is called triangulation.
All numerical methods for summarizing a set of data are descriptive statistics. The mean is the arithmetic average, determined by adding the scores and dividing by the sum of the number of scores. The median is the middle most score, determined by ranking scores from highest and lowest and noting the one in the exact centre. For some comparisons, the variability of a set of numbers must also be determined. When the scores cluster close to the mean, they have a low variability. Standard deviation is the most common measure of variability. The further the scores from the mean, the greater the standard deviation.
When both variables of a correlational study are measured numerically, a statistic called the correlation coefficient can be determined. A formula produces a result between +1.00 and -1.00. The direction of the correlation might be positive (+: the increase of one variable causes the other to increase) or negative (-: the increase of one variable causes the other to decrease). To visualize the relationship between variables, a scatter plot may be used. From this, you can see how strong the correlation is and in what direction it goes.
Inferential statistics are necessary to determine how confident a research can be in inferring a general conclusion from data.
When two means are compared, p is the probability that the difference is as big or bigger than if the independent variable had no effect and the result was a matter of chance. When the p is less than 0.05%, the results can be considered statistically significant. Statistic significance tests have three components:
Size of effect— If an effect is large, chances are it is also significant.
Number of subjects or observations in the study—the larger the sample, the more accurately the observed mean will reflect the true mean.
Variability of data within the group—when the group means are compared and an index of variability is created, it can be determined how different the scores are from one another. The higher the variability, the higher the possible randomness of the result.
Ideally, bias and error should be minimized. Error is the random variability in results and is inevitable in most research. Error can often be measured and corrected for. Bias includes non-random effects caused by extraneous factors. Bias is hard to identify and cannot be corrected for with statistics.
If members of one group are chosen differently than those in another group, the sample might be considered biased. A sample is biased when not representative of the larger population it is supposed to describe. Random assignment is a method for counteracting sample bias.
When a test can be repeated with a particular subject in a particular set of conditions and produce similar results, it is considered reliable. If the scores are greatly affected by the whims of the subjects, the test has low reliability. Validity is the extent to which a test measures what it claims to measure. If a test lacks validity, it is likely biased. Face validity is how valid a test seems to be, according to common sense. Criterion validity is determined by correlating scores with a more direct index of the desired characteristic of study.
External validity looks at whether a sample is generalizable to the population and whether the setting of the study is generalizable to the real world. When your participants are volunteers, you may get other results than when your participants are non-volunteers (psychology students).
Researchers have wishes and expectations that might affect their behaviour and observations – this is the observer-expectancy bias. If a desire is communicated unintentionally, the subject might pick up on this and behave according to expectation.
An example of observer-expectency can be seen in the development of “facilitated communication”, in which a facilitator would help autistic people type by holding their hands up to a keyboard. At first, it seemed as if the autistic children were truly communicating. However, further research discovered that the facilitators were subconsciously influencing he movements of the autistic child’s finger by the subtle motions of their hand, and the resulting text was not communication controlled by the autistic child.
Observer expectations might not only influence the subjects’ behaviour, but also the observer’s observations. To prevent this, the observer can be kept blind (uninformed) about the aspects of the study that might lead them to form biased expectations. Ensuring that they don’t know which group in a between-group study has been exposed to an altered independent variable can keep observer-expectancy to a minimum.
When subjects have expectations, the results of an experiment can be biased. A double blind keeps both the observer AND the subject uninformed of whether they are or are not in a control group, receiving a placebo.
Three ethical issues must be considered when conducting research with humans:
Right to privacy: Informed consent should be obtained before the subjects take part, and they should be informed that they do not have to share information they don’t want to share.
Possibility of harm: If a study involves discomfort or harm, the psychologist must determine whether the same hypothesis can be equally tested in a harmless experiment. Subjects must also be reassured that they can quit at any time.
Deception: In some experiments, the independent variable involves a lie. Some believe that deception is intrinsically unethical and undermines truly informed consent. Others justify deception as necessary for the study of certain psychological processes.
JoHo donateurs ontvangen 10% korting op de scriptiebegeleiding van ScriptieMaster.
word JoHo donateur of abonnee voor online toegang
Korte advieswijzer JoHo services
JoHo donateur (= €5 per kalenderjaar):
- Voor wie JoHo wil steunen
- Voor wie toegang wil tot de (service)abonnementen
- Voor wie van basiskortingen gebruik wil maken
JoHo donateur met service-abonnement I (= €5 + €10 per kalenderjaar): upgrade naar abonnement I
- Voor wie toegang wil tot de online JoHo's en boeksamenvattingen
- Voor wie gebruik wil maken van de vacatureservice en bijbehorende keuzehulp & advieswijzers
- Voor wie gebruik wil maken van keuzehulp en advies wil bij werk in het buitenland, lange reizen, vrijwilligerswerk, stages en studie in het buitenland
JoHo donateur met service-abonnement II (= €5 + €20 per kalenderjaar): upgraden naar abonnement II
- Voor wie online volledig gebruik wil maken van alle JoHo's en boeksamenvattingen voor alle fases van een studie, met toegang tot alle online HBO & WO boeksamenvattingen en andere studiehulp
- Voor wie extra kortingen wil op de JoHo's en boeksamenvattingen
- Voor wie extra kortingen wil op (reis)artikelen en services
- Voor wie gebruik wil maken van de emigratie- en expatservice
JoHo donateur met service-abonnement III (= €5 + €40 per kalenderjaar):
- Voor wie gebruik wil maken van een cv-check, persoonlijke adviesservices & de hoogste kortingen op artikelen, samenvattingen en services
JoHo donateur met een doorlopende verzekering
- Sluit je via JoHo een jaarlijks doorlopende verzekering af, dan kan je gedurende de looptijd van je verzekering gebruikmaken van de voordelen van service-abonnement III: hoge kortingen + volledig online toegang + alle extra services. Lees meer
Je bent al donateur, maar je hebt geen toegang
Indien het een adviespagina, keuzehulppagina of een (exclusieve) samenvatting betreft, dan heb je een (service)abonnement nodig om toegang te krijgen tot deze pagina. Lees hierboven meer over welk abonnement je kunt afsluiten.
Wil je tijdelijk extra toegang, bijvoorbeeld tot de Exclusive samenvattingen of keuzekits en advieswijzers, sluit dan als donateur een maandabonnement met volledige online toegang af. Lees hier meer over de voordelen en services bij dit tijdelijke (service)abonnement.
Samenvattingen & Studiehulp: Mogelijk heb je als JoHo donateur heb je vorig studiejaar, ter kennismaking, gebruik kunnen maken van de de toegang tot de online JoHo's (samenvattingen en studiehulp) voor de propedeusevakken. Dit studiejaar kan je gebruik blijven van de JoHo's door te 'upgraden' naar abonnement I of abonnement II en een voordelig JoHo (service)abonnement af te sluiten. Daarnaast kan je ter waarde van de kosten van je abonnement gratis JoHo's komen afhalen in de JoHo support centers!
Lees hier de antwoorden op de meest gestelde vragen.
Meer en uitgebreide advieswijzers
voor samenvattingen en stages - voor vacatures en sollicitaties - voor reizen en backpacken - voor vrijwilligerswerk en duurzaamheid - voor emigratie en lang verblijf in het buitenland- voor samenwerken met JoHo
Steun JoHo en steun jezelf door JoHo donateur te worden
Bezoek gratis de Nederlandse carrièredagen in de Jaarbeurs Utrecht!
JoHo is een ontwikkelingsorganisatie met wereldwijd winkels en websites,
waar mensen en organisaties worden gestimuleerd en geholpen bij talentontwikkeling en internationale samenwerking
|Samenwerken & Adverteren||World Supporter platform||World Summaries platform||Expat platform||World Activity Platform||Emigratie platform||Jouw account pagina|
|JoHo||JoHo & Leren||JoHo & Werken||JoHo & Reizen||JoHo & Emigreren||You & JoHo|