BulletPointsummary of Research Methods in Psychology: Evaluating a World of Information by Morling - 4th edition

What is the psychological way of thinking? - Bulletpoints 1

• Some psychology students would like to become a researcher or a scientists. They are called producers of research. Other psychology students would rather not become a researcher or work in a laboratory, but like to read research in order to apply it to their job, hobbies or relationships. They are known as consumers of research. In practice, psychologists are both producers and consumers of research. Producers and consumers share a love for empiricism: they want to answer psychological questions through direct and formal observation.
• Empiricism, the empirical method or empirical research means that evidence from the senses or instruments that help the senses (such as photographs, questionnaires, measuring scales, thermometers or timers) are used to draw conclusions. Empiricists want to be systematic and rigorous in their research, and want to make sure other researchers can verify or debunk their conclusions independently.
• The theory-data cycle means that scientists collect data to test their theories, and change or adapt them. This happens as follows: first, a scientist asks a question about a phenomenon they want to be able to explain with their theory. Then, they make a prediction or predictions about how the phenomenon could be explained. They then test these predictions by collecting data. Based on the data they determine whether they should adapt or change their theory, and whether to ask follow-up questions.
• Applied research means that researchers do a study based on a real life practical problem. Basic research aims to increase the overall knowledge we have about certain topics, without focusing on a practical problem specifically. Translational research makes use of the knowledge from basic research to develop and test treatments. It forms the bridge between basic and applied research.

What are the different sources of information and why is scientific research the best source? - Bulletpoints 2

• A comparison group makes sure that we can look at what happens in a situation in which the variable that is being studies is present, versus a situation in which this variable is not present.
• When research results do not fit our own experiences we often conclude that the research must be wrong. However, research is probablistic. This means that research results do not always explain all cases, but try to explain the biggest proportion as they possibly can. Research suggests a strong possibility or high probability that something will happen, but this prediction is not perfect.
• Empirical articles write about certain research results for the first time. They discuss the method, statistical tests and results of the research. Review articles integrate all published research on one topic and summarize it. Sometimes review articles use meta-analysis, in which the results of different studies are combined and proportionally influence an effect-size: a number that states the size or strength of a relationship between different variables. Both empirical and review articles are peer reviewed.

Which types of claims and validity can consumers of research interrogate? - Bulletpoints 3

• A variable is something that can vary, so it needs to have at least two levels. A constant is something that can vary, but has one level in a study. In research, every variable is measured or manipulated. A measured variable is a variable whose levels are observed and written down. A manipulated variable is a variable a researcher manipulates and thus controls, usually by assigning people that take part in a study to different levels of that variable.
• Conceptual variables or constructs are abstract concepts that basically name the concepts which are being studied. An example of this is "perseverence." To make hypotheses about these conceptual variables, theoretical definitions must be ascribed. These are known as conceptual definitions. An example would be: "whether someone can keep going when things go wrong." This conceptual definition is then operationalized. This means that a concept is translated to a variable that can be measured or manipulated. This new variable is also known as an operational variable.
• Frequency claims describe the frequency of a variable, or the rate in which a variable is present: how often something happens. Association claims state something about the level of one variable being associated with the level of another variable. Causal claims state that one variable is the cause of a change in the other variable.
• Construct validity shows how well a conceptual variable was operationalized. External validity is about how well the results of a study generalize to the real world. Statistical validity looks at to what extent the statistical conclusions of a study are precise, reasonable and replicable.

What are the ethical guidelines for psychological research? - Bulletpoints 4

• The principle of respect for persons: means that participants should be treated like autonomous agents. They should decide for themselves whether to take part in research or not. They should be able to give informed consent. This means that participants are provided information about the goal of the research, the risks and benefits and then decide based on this whether or not they want to take part in the study.
• The principle of beneficence: means researchers need to know beforehand whether participants are at risk during the research and what the benefits could be. The risks and benefits for the population should also be checked.
• The principle of justice: looks for a balance between the people that partake in the study and the people who get the benefits from it. Researchers should know to what extent the participants that take part in the study are an accurate representation of the group that gets the benefits from it.
• The five general ethical principles of the APA are: respect, humanity, justice, integrity and fidelity and responsibility.

What are good measures in psychology? - Bulletpoints 5

• When a variable is operationalized, a construct is turned into a measured or manipulated variable. Psychological research always requires two definitions of a variable. The first is a conceptual definition or construct: the definition that the researcher gives to the variable on a theoretical level. The operational definition reflects how a researcher decided to measure or manipulate a variable.
• Operationalized variables are classified primarily as categorical or quantitative. A categorical variable or nominal variable has categories as levels. A quantitative variable or continuous variable is coded with numbers that do have a numerical meaning, for example weight or height. Here the numbers do indicate that someone scores higher or lower than another person.
• We can determine reliability in three ways: test-retest reliability, interrater reliability and internal reliability.
• The correlation coefficient or r is a number that represents how close different measurements or points are to a line drawn on a scatterplot. It shows the direction and strength of a relationship between variables.
• The types of validity we can check are: face validity, content validity, criterion validity, convergent validity and discriminant or divergent validity.

How do we use surveys and observations? - Bulletpoints 6

• The way we can ask questions are as an open question, a closed question, a likert scale or a semantic differential.
• Subjects can sometimes give less accurate answers. They don't always do this intentionally. Sometimes there are response sets or non-differentiation: these are a kind of 'shortcut' that people take when answering survey questions. These can reduce the construct validity because respondents often adopt response patterns in response sets instead of expressing their opinion.
• When a researcher observes the behaviour of animals or humans and notes it systematically, then he/she is conducting observational research.  With observational research, researchers work very carefully to make sure the observations are accurate and valid. Observations have a good construct validity when they can avoid the following three problems: observer bias, observer effects and reactivity.

How do we estimate the frequencies of behaviors and attitudes? - Bulletpoints 7

• When you test external validity, you wonder whether the results of a certain research can be generalized to a bigger population. he external validity is really important for frequency claims. You wonder whether the results you found for your participants can be found in the entire population. Does your sample represent the entire population of interest? External validity does not just look at a sample, but also at a setting.
• A population can be seen as a whole set of people or products a researcher is interested in. A sample is a smaller set from that population.
• There are two ways in which a sample may be biased. First, researchers may have only included people they could easily contact in a sample. This is also known as convenience sampling. It is also possible that researchers only include people in the sample who voluntarily register to participate. This is called self-selection. Both convenience sampling and self-selection can threaten external validity because people who are easier to reach or offer themselves for participation are not necessarily representative of the population of interest.
• If researchers really want to have a representative sample, they could best use a probability sample. Probability sampling is also called random sampling. This means that every member of the population of interest has an equal chance to be chosen for the sample.
• When researchers only want to ask people from a certain subgroup and these people are not chosen in a random way, they are using purposive sampling. Another form of purposive sampling is snowball sampling. The participant is asked to ask a couple of relatives or acquaintances he/she knows to participate in the research. Of course, this is not a representative way to sample people, because people ask others from their social networks and that is not a random way. In quota sampling, a researcher identifies a subpopulation and he chooses how big every subpopulation in the sample will be. Then, he chooses the people from this population in a non-random way.

What is bivariate correlational research? - Bulletpoints 8

• Association claims are claims that describe the relationship between two measured variables. A bivariate correlation is also called a bivariate association between two variables. In order to study an association, we have to study both variables within the same group of people.
• The first thing that determines the statistical validity of a study is how strong the relationship between the variables is. This is shown as an effect size. An effect size thus describes how strong the relationship between two or more variables is. To describe an effect size, the terms weak, medium and strong are used.
• An outlier is an extreme score that is not in line with the other scores. Depending on the way in which the score stands out, it can have a major influence on the correlation between the variables. An outlier can make a correlation appear stronger or weaker than it is. In a bivariate correlation, outliers can be especially problematic if there are extreme scores on both variables.
• A range restriction means that the full range of scores on one of the variables is not covered, making a correlation appear smaller than it actually is.
• A curvilinear association means that the relationship between two variables is not a straight line. We speak of this, for example, when a relationship between two variables is first negative and then suddenly becomes positive after a certain point. This greatly affects the r-value, which represents a straight best-fit line through the data.

What is multivariate correlational research? - Bulletpoints 9

• Longitudinal designs can show temporal precedence by measuring the same variables in the same person in different times. Longitudinal designs are often used in developmental psychology to study the changes in certain characteristics of humans.
• The different correlations of multivariate correlational research are: cross-sectional correlations, autocorrelations and cross-lag correlations.
• Multiple regression or multivariate regression is a statistical technique where we can exclude certain third variables. As a result, they no longer exert any influence on the relationship between two particular variables and this allows us to guarantee the internal validity of a study. This means that multiple regression controls for the influence of a possible third variable on the relationship between two others. In order to perform multiple regression, you should always measure a number of other variables in addition to the two main variables that you want to investigate so that you can investigate their mutual relationships. This is called a multivariate correlational study. n a multiple regression, therefore, three or more variables are considered. The variable that the researcher is most interested in is called the criterion variable or the dependent variable. The other variables in a regression analysis are called the predictor variables or the independent variables.
• Parsimony is the degree to which a good scientific theory can provide the simplest explanation for a phenomenon. In causal claims, parsimony refers to the simplest explanation for a pattern in your data.

How can causal claims be evaluated with the help of experiments? - Bulletpoints 10

• In psychology, doing an experiment means that a researcher manipulates at least one variable and measures another variable. Experiments can take place in a lab or elsewhere, as long as at least one variable can be manipulated and one variable can be measured. A manipulated variable is a variable that the researchers try to control, for example by assigning participants to a certain level of the variable. Measured variables are recorded behaviors or attitudes. This is often done through self-reporting, behavioral observation or physiological measurements. The variable being manipulated is also known as the independent variable while the measured variable is also known as the dependent variable. The different levels of the manipulated/independent variable to which researchers attribute participants are also referred to as conditions.
• A control group is a level of the independent variable that shows ‘no treatment’ or a neutral condition. When a researcher has a control group, the other conditions in which the level of the independent variable is varied are called treatment groups. If we expose a group to a treatment, but this treatment does not do or change anything we call this a placebo group or placebo control group.
• In an independent-groups design, different groups of participants are assigned to different levels of the independent variable. This is also called a between-group design. In a within-groups design (also called a within-subjects design), there is just one group of participants and every person is be exposed to every level of the independent variable.

Where and how can we determine the influence of confounding and obscuring factors? - Bulletpoints 11

• A one-group, pretest/posttest design is a design in which a researcher recruits a group of participants, measures them on a particular variable in a pretest, then exposes them to a treatment or intervention, and then measures them on a posttest.
• The six threats to the internal validity for a one-group, pretest/posttest design are: maturation threat, history threat, regression threat, attrition threat, testing threat and instrumentation threat. Combined threats can also happen.
• The three threats that occur to internal validity in each study are: observer bias, demand characteristics and placebo effects.
• A null effect means that the independent variable has had no influence on the dependent variable. There appears to be no significant covariance between the two.
• Power is an aspect of statistical validity; the probability that a study will show an accurate result if the independent variable actually has an effect on the dependent variable.

How do we deal with experiments that have more than one independent variable? - Bulletpoints 12

• An interaction effect looks at if the effect of the original independent variable depends on the level of other independent variable.
• A factorial design is a design with two or more independent variables (called factors). Usually, the two independent variables are crossed. That means that researchers test every possible combination of the independent variable.
• A moderator is a variable that influences the relationship between the independent variable and the dependent variable. A moderator results in an interaction: the effect of one independent variable depends (or is moderated by) the level of another variable. 
• The different types of factorial design are: independent-groups, within groups or mixed factorial design.

What are quasi-experiments and small-N designs? - Bulletpoints 13

• A quasi-experiment differs from actual experiment in the sense of control. In a quasi-experiment researchers do not have full control over the conditions, because participants are not randomly assigned. Because researchers usually have no control over the independent variable, it is also called a quasi-independent variable.
• The types of quasi-experiments are: nonequivalent control group pretest/posttest design, nonequivalent control group posttest-only design, interrupted time-series design and nonequivalent conrol group interrupted time-series design.
• N is the number of subjects in a sample. It is more important for external validity to select a sample in the right way than to include many subjects in the sample. When researchers use a small-N design, instead of getting little information from a large sample, they get a lot of information from a small sample. They can even look at one animal or person in a single-N design.
• The three different small-N designs are: stable-baseline design, multiple-baseline design and a reversal design.

Can we apply the results of a study to the real world? - Bulletpoints 14

• Scientists should always ask themselves whether the results of their research are replicable. If something is replicable, it means that the findings would remain the same if the same research was done again. Replicability gives a study credibility. Often, researchers replicate their results before their findings are published.
• The types of replication are direct replications, conceptual replications and replication-plus-extension studies.
• Scientific literature consists of a series of related studies conducted by different researchers and testing similar variables. Sometimes researchers collect all studies on a specific topic and they turn it into a review article. One way of writing this review article is by reciting all the findings. Another way is to make a mathematical summary of the scientific literature. That is called a meta-analysis.This is a way to mathematically average the results of both published and unpublished studies that have examined the same variable in order to draw a conclusion about the an effect as a whole.
• Questionable research practices are things in science that go against Merton's standards.
• In the theory-test modus, researchers only want to test an association that might support their theory. In that case, it is more important to test internal validity than external validity. In a generalizable modus, these psychologists want to generalize the results of their samples to a bigger population.