The disease model states that problems are symptoms of a small set of underlying disorders. This explains observable clinical symptoms by a small set of latent variables (e.g. depression). A network is a set of elements (nodes) connected through a set of relations. In network models, disorders are conceptualized as systems of causally connected symptoms rather than effects of a latent disorder.

Mental disorders cannot be identified independently of their symptoms. In medicine, the medical condition can be separated from the symptoms. In psychology, this is not possible. In order to separate this, it must be possible that a person has symptoms without the disorder (e.g. depression without feeling down is not possible). In mental disorders, it is likely that there is symptom-symptom causation. One symptom causes another symptom and this leads to a mental disorder.

With network systems, it might be unclear where one disorder starts and another stops. The boundaries between disorders become unclear. Network models might change treatment, as the treatment is then no longer aimed at the disorder but rather at the symptoms and the causal relationship between the symptoms.

Networks in psychopathology can be created by using data on symptom endorsement frequencies (e.g. looking at correlations between symptoms) (1), assess the relationship between symptoms rated by clinicians and patients (2) and use the information in the diagnostic systems (3).

In networks, any node can reach another node in only a few steps. This is called the small world property. The DSM attempts to be neutral, theoretically, but makes claims about causal relationships between the disorders.

Asking experts on how nodes are related (e.g. clinicians and symptoms of a disorder) is called perceived causal relations scaling

Extended psychopathology systems refers to network systems in which the network is not isolated in a single individual but spans across multiple individuals. This would mean that one symptom in one person could cause a symptom in another person. These networks can be used to review what the interaction is between symptoms of different people in different social situations.

Association networks show what the strength of the correlations between symptoms is. This gives an indication of different disorders, as the symptoms in disorder A are more correlated with the other symptoms in disorder A than with the symptoms of disorder B.

A partial correlation network, also called a concentration network, shows the partial correlations between symptoms. This can be used to be a bit more certain about the causal relationship of two nodes as it rules out some third-variable explanations. Concentration graphs can be used to assess which pathways between symptoms appear common in a disorder.

Association and concentration graphs provide information about the causal relationship between nodes but it does not provide information about the causal direction of the network. Directed networks give information about the causal relationship between nodes. This is usually represented in a DAG. In order to generate statements

The item response theory (latent trait theory) provides a way to model the probability that a person with X ability will be able to perform at level of Y. It models the probability that a person with X amount of a personality trait will exhibit Y amount of that trait on a test that is supposed to measure is. This theory focusses on the relationship between a testtaker’s response to an individual test item and that testtaker’s standing on the construct being measured.

Discrimination signifies the degree to which an item differentiates among people with higher or lower levels of the trait. Items can be given different weight in the item response theory. In classical test theory, there are no assumptions about the frequency distribution of test scores.

There are several assumptions of the item response theory:

1. Unidimensionality
   This assumption states that the set of items measures a single continuous latent construct. This assumption does not neglect minor dimensions, although assumes one dominant dimension underlying the structure.
2. Local independence
   This assumption states that there is a systematic relationship between all of the test items and this relationship has to do with the level of a person on the construct of interest. If this assumption is met, then the differences in responses to items are reflective of differences in the underlying trait or ability.
3. Monotonicity
   This assumption states that the probability of endorsing or selecting an item response indicative of higher levels of the construct should increase if as the level of the underlying construct increases.

Local dependence refers to the fact that items can be dependent on another factor than what the test as a whole is measuring. Locally dependent items have higher inter-item correlations and it may be controlled for by combining the responses to a set of locally dependent items into a separate subscale within the test. The theta level refers to the level of the underlying construct.

The probabilistic relationship between a testtaker’s response to a test item and that testtaker’s level on the latent construct being measured can be expressed in the item characteristic curve (ICC).

IRT enables test users to better understand the range of the underlying construct for which an item is most useful in discriminating among groups of testtakers. This can be done using the information function.

Information refers to the precision of measurement.

Items with low information prompt the test developer to consider the possibility that the content of the item does not match the construct measured by the other items (1), the item is poorly worded (2), the item is too complex (3), the placement of the item in the test is out of context (4) or cultural factors may be operating to weaken the item’s ability to discriminate between groups (5).

Qualitative research refers to the collection and analysis of non-numerical data through a psychological lens in order to provide rich descriptions and possibly explanations of people’s meaning-making, how they make sense of the world and how they experience particular events.

Epistemology refers to the theory of knowledge regarding what we can know and how we can know. Ontology refers to the assumptions made about the nature of being, existence or reality. Different research approaches are associated with different epistemologies.

Positivism holds that there is a direct correspondence between the state in the world and our perceptions through our senses, provided that our perception is now skewed by factors that could damage that correspondence (e.g. interest in a topic). Empiricism states that our knowledge must arise from the collection and categorization of our sense perceptions of the world. Hypothetico-deductivism states that theories should be exposed to attempts of falsifications, rather than attempts of verification.

The classical scientific method assumed that reality exists independently of the observer and that reality can be observed through research. It assumes that any existing psychological dimension could be measured with precision.

‘Small q’ qualitative research is a structured form of content analysis, which categorizes and quantifies qualitative data systematically. ‘Big Q’ qualitative research refers to the use of qualitative techniques within a qualitative paradigm which rejects notions of objective reality or universal truth.

Nomothetic research seeks generalizable findings that uncover laws to explain objective phenomena and idiographic research seeks to examine individual cases in detail to understand an outcome. Phenomenological methods focus on obtaining detailed descriptions of experience as understood by those who have that experience in order to discern its essence.

Critical realism states that reality exists independent of the observer, although we cannot know that reality with certainty. Social constructionism has a critical stance towards assumptions about the world. It states that the way we understand the world and ourselves are built up through social processes. This is not fixed. Relativism states that reality is dependent on the ways we come to know it.  

Reflexivity refers to the acknowledgement by the researcher of the role played by their interpretative framework in creating their analytic account.

Sensitivity to context refers to whether the context of the theory is made clear. Commitment refers to prolonged engagement with the research topic. Rigour refers to the completeness of the data collection and analysis. Coherence refers to the quality of the research narrative and the fit between the research question and the adopted philosophical perspective. Impact and importance refers to the theoretical, practical and socio-cultural impact of the study.

There are several evaluative criteria for qualitative research:

Probabilities are long-run relative frequencies for the collective, rather than an individual. Probabilities do not apply to theories, as individual theories are not collectives. Therefore, the null hypothesis cannot be assigned a probability. A p-value does not indicate the probability of the null hypothesis being true.

Power or a p-value is not necessary in Bayesian statistics, as a degree of plausibility can be assigned to theories and the data tells us how to adjust these plausibilities. It is only needed to determine a factor by which we should change the probability of different theories given the data.

The probability of a hypothesis being true is the prior probability (P(H)). The probability of a hypothesis given the data is the posterior probability (P(H|D)). The probability of obtaining the exact data given the hypothesis is the likelihood (P(D|H)). Therefore, the posterior probability is the likelihood times the prior probability.

The likelihood principle states that all information relevant to inference contained in data is provided by the likelihood. In a distribution, the p-value is the area under the curve at a certain point. The likelihood is the height of the distribution at a certain point.

The p-value is influenced by the stopping rule (1), whether or not the test is post-hoc (2) and how many other tests have been conducted (3). These things do not influence the likelihood.

The Bayes factor is the ratio of the likelihoods. The Bayes factor is driven to 0 if the null hypothesis is true, whereas the p-values fluctuate randomly if the null hypothesis is true and data-collection continues. The Bayes factor is slowly driven towards the ‘truth’. Therefore, the Bayes factor gives a notion of sensitivity. It distinguishes evidence that there is no relevant effect from no evidence of a relevant effect. It can be used to determine the practical significance of an effect.

Adjusting conclusions according to when the hypothesis was thought of would introduce irrelevancies in inference and therefore, the timing of the hypothesis is irrelevant in Bayesian statistics. In assessing evidence for or against a theory, all relevant evidence should be taken into account and the evidence should not be cherry picked.

Rationality refers to having sufficient justification for one’s beliefs. Critical rationalism is a matter of having one’s beliefs subjected to critical scrutiny. Irrational beliefs are beliefs not subjected to sufficient criticism.

It is possible to have a uniform (1), normal (2) and half-normal (3) distribution. In a uniform distribution, all values are equally likely. In a normal distribution, one value is most likely given the theory and a half-normal distribution is a normal distribution centred on zero with only one tail. It predicts a theory into one direction but smaller effects are more likely than larger effects.

There are several weaknesses of the Bayesian approach:

1. Bayesian analyses force people to specify predictions in detail
2. Bayesian analyses do not

The problem of causality is difficult in developmental psychology, as many questions of that field regard factors that a person cannot be randomly assigned to (e.g. single parent family). Causal inference refers to the study and measurement of cause-and-effect relationships outside of random assignment.

In the current situation in developmental psychology, it is unclear among researchers whether causality can be implied and why. Causal inferences are necessary for the goals of developmental psychology because causal inferences can improve the lives of people (1), can help distinguish between associations and causal claims for laypeople (2) and causal thinking is unavoidable (3).

The directed acyclic graph (DAG) is a tool which is useful in moving from associations to causal relationships. It is particularly useful in identifying covariates and understanding the anticipated consequences of incorporating these variables.

The DAG is a symbolic representation of dependencies among variables. The causal Markov assumption states that the absence of a path (in the DAG) implies the absence of a relationship. In the DAG, models that represent data with fewer links are preferred to the more complex (parsimony). If two variables are simultaneously determined, the DAG could incorporate this possibility by treating the two as reflecting a common cause.

Variables (in the DAG) can be related in three ways:

1. Z is a common cause of X and Y
   In this case, Z needs to be controlled for.
2. Z is a common effect of X and Y
   This is a collider. Conditioning on a collider creates a spurious relationship between X and Y. This relationship can suppress or inflate a true causal effect.
3. Z mediates the effect of X on Y

    

There are four procedures to present the implications of a correlation in terms of our ability to use the correlations to make successful predictions:

1. Binomial effect size display (dichotomous)
   This illustrates the practical consequences of using correlations to make decisions. It can show how many successful and unsuccessful predictions can be made on the basis of a correlation. It uses the following formula:
   
2. Binomial effect size display can be used to translate a validity correlation into an intuitive framework. However, it frames the situation in terms of an ‘equal proportions’ situation.
3. Taylor-Russell tables (dichotomous)
   These tables inform selection decisions and provide a probability that a prediction will result in a successful performance on a criterion. The size of the validity coefficient (1), selection proportion (2) and the base rate (3) are required for the tables.
4. Utility analysis
   This frames validity in terms of a cost-benefit analysis of test use.
5. Analysis of test sensitivity and test specificity
   A test is evaluated in terms of its ability to produce correct identifications of a categorical difference. This is useful for tests that are designed to detect a categorical difference.

Validity correlations can be evaluated in the context of a particular area of research or application.

A nomological network refers to the interconnections between a construct and other related construct. There are several methods to evaluate the degree to which measures show convergent and discriminate associations:

1. Focusses associations
   This method focusses on a few highly relevant criterion variables. This can make use of validity generalization.
2. Sets of correlations
   This method focusses on a broad range of criterion variables and computes the correlations between the test and many criterion variables. The degree to which the pattern of correlations ‘makes sense’ given the conceptual meaning of the construct is evaluated.
3. Multitrait-multimethod matrices
   This method obtains measures of several traits, each measured through several methods. The purpose is to set clear guidelines for evaluating convergent and discriminant validity evidence. This is done by evaluating trait variance and method variance. Evidence of convergent validity is represented by monotrait-heteromethod correlations.

The correlations between measures are called validity coefficients. Validity generalization is a process of evaluating a test’s validity coefficients across a large set of studies. Validity generalization studies are intended to evaluate the predictive utility of test’s scores across a range of settings, times and situations. These studies can reveal the general level of predictive validity (1), reveal the degree of variability among the smaller individual studies (2) and it can reveal the source of the variability among studies (3).

Good science requires statistical tools and informed judgement about what model to construct, what hypotheses to test and what tools to use.

There is no universal method of scientific inference, but, rather, a toolbox of useful scientific methods. Besides that, the danger of Bayesian statistics is that this will become a new universal method of statistics. Lastly, statistical methods are not simply applied to a discipline, they change the discipline itself.

In natural sciences, the probabilistic revolution shaped theorizing. In social sciences, it led to scientists mechanizing scientists’ inferences. Inference revolution refers to the idea that inference from sample to population was considered the most important part of research. This revolution led to a dismissive attitude towards replication.

There are three meanings of significance:

1. Mere convention
   This means that it is convenient for researchers to use 5% as a standard level of significance.
2. Alpha level
   This means that significance refers to the long-term relative frequency of making a type-I error.
3. Exact level of significance
   This is the exact level of significance and is used in null hypothesis testing using a nil and not a null of zero difference.

There are three interpretations of probability:

1. A relative frequency
   This is a long-term relative frequency.
2. Propensity
   This is the physical design of an object (e.g. dice)
3. Reasonable degree of subjective belief
   This is the degree an individual believes in somethings.

Bayesian statistics should not be used in an automatic way, like frequentism. Objections to the use of Bayes rule are that frequency-based prior probabilities do not exist (1), that the set of hypotheses needed for the prior probability distribution is not known (2) and that researchers’ introspection does not confirm the calculation of probabilities (3).

Fishing expeditions refers to the idea that hypothesis finding is the same as hypothesis testing, characterised by using a lot of p-values in a research article.

Simpson’s paradox states that the direction of an association at the population-level may be reversed within subgroups of that population. Inadequate attention to the Simpson’s paradox may lead to faulty inferences. The Simpson’s paradox can arise because of differences in proportions on subgroup levels compared to population levels. It also states that a pattern (association) does not need to hold within a subgroup.

The paradox is related to a lot of things, including causal inference. A generalized conclusion (e.g. extraversion causes party-going) might hold for the general population, but does not mean that this inference can be drawn at the individual level. A correlation across the population does not need to hold in an individual over time.

In order to deal with Simpson’s paradox, the situations in which the paradox occurs frequently have to be assessed. There are several steps in preventing Simpson’s paradox:

1. Consider when it occurs.
2. Explicitly propose a mechanism, determining at which level it is presumed to operate.
3. Assess whether the explanatory level of data collection aligns with the explanatory level of the proposed mechanism.
4. Conduct an experiment to assess the association between variables.

In the absence of strong top-down knowledge, people are more likely to make false inferences based on Simpson’s paradox.

One way of avoiding the null hypothesis testing ritual in science is to increase the precision of theories by casting them as formal models. Rituals can be characterized by a repetition of the same action (1), fixations on special features (2), anxieties about punishment for rule violation (3) and wishful thinking (4). The null hypothesis testing ritual is mainly maintained because many psychological theories are too weak to make precise predictions besides the direction of the effect.

A model is a simplified representation of the world that aims to explain observed data. It specifies a theory’s predictions. Modelling is especially suited for basic and applied research about the cognitive system. There are four advantages of formally specifying the theories as models:

1. Designing strong tests of theories
   Modelling theories leads to being able to make quantitative predictions about a theory, which then leads to comparable, competing predictions between theories which allows for comparison and testing of theories.
2. Sharpening research questions
   Null hypothesis testing allows for vague descriptions of theories and specifying the theories as models requires more precise research questions. These vague descriptions make theories difficult to test and sharpening the research questions makes it easier to test the theories.
3. Going beyond linear theories
   Null hypothesis testing is especially applicable to simple hypotheses. The statistical tools available are used to create theories, mostly linear theories and by specifying the theory as a model, this is not necessary anymore.
4. Using more externally valid designs to study real-world questions
   Modelling can lead to more externally valid designs, as confounds are not eliminated in the analysis, but built into the model.

Goodness-of-fit measures cannot make the distinction between variation in the data as a result of noise or as a result of the psychological process of interest. A model can end up overfitting the data, capturing the variance of the psychological process of interest and variance as a result of random error. The ability of a model to predict new data is the generalizability. The complexity of a model refers to a model’s inherent flexibility that enables to fit diverse patterns of data. The complexity of a model is related to the degree to which a model is susceptible to overfitting. The number of free parameters (1) and how parameters are combined in the model (2) contribute to the model’s complexity.

Increased complexity makes a model more likely to overfit while the generalizability to new data decreases. Increased complexity can also lead to better generalizability of the data, but only if the model is complex enough and not too complex. A good fit to current data does not predict a good fit to other data.

The irrelevant specification problem refers to the difficulty bridging the gap between description of theories and formal implementations. This can lead to unintended discrepancies between theories and their formal counterparts. The Bonari paradox refers to when models become more complex and

Before the existence of the IAT, indirect measures of prejudice were developed in order to overcome response bias and psychologists began to examine automatic processes that may contribute to contemporary forms of prejudice. After the existence of the IAT, implicit prejudice became the same thing as widespread unconscious prejudices that are more difficult to spot and regularly infect intergroup interactions.

The IAT has been used throughout different areas of society and is a very popular mean of describing implicit prejudice. Prejudice extends beyond negative or positive associations with an attitude object to include motivational and affective reactions to in-group and out-group members. IAT does not have a strong predictive validity. The IAT score is a poor predictor of discriminating behaviour.

There are no guidelines for how to interpret the scores on the IAT. This is referred to as the score interpretation problem. The test scores are dependent on arbitrary thresholds and it is not possible to link them to behaviour outcomes.

The focus of the IAT on implicit gender stereotypes is (not implicit sexism) is problematic because implicit measures of gender stereotypes are not a good predictor of discriminatory behaviour (1), only a very limited set of implicit gender stereotypes has been examined (2) and no explanation is provided about how conflicts between automatic evaluative associations and automatic semantic associations are resolved (3).

Individuating information, getting personal information about a certain group, exerts effects to counter explicit biases. It does the same with regard to implicit biases.

Subjective evaluation criteria are not associated with discrimination. Therefore, the solution that only objective measures must be used in decision making to counter (implicit) bias is unnecessary. This is referred to as the subjective judgement problem.

Confounding bias occurs when a variable influences both who is selected for the treatment and the outcome of the experiment. If a possible confounding variable is known, it is possible to control for the possible confounding variable. Researchers tend to control for all possible variables, which leaves the possibility of controlling for the thing you are trying to measure (e.g. controlling for mediators).

Confounding needs a causal solution, not a statistical one and causal diagrams provide a complete and systematic way of finding that solution. If all the confounders are controlled for, a causal claim can be made. However, it is not always sure whether all confounders are controlled for.

Randomization has two clear benefits. It eliminates confounder bias and it enables the researcher to quantify his uncertainty. Randomization eliminates confounders without introducing new confounders. In a non-randomized study, confounders must be eliminated by controlling for them, although it is not always possible to know all the possible confounders.

It is not always possible to conduct a randomized controlled experiment because of ethical, practical or other constraints. Causal estimates of observational studies can provide with provisional causality. This is causality contingent upon the set of assumptions that the causal diagram advertises.

Confounding stands for the discrepancy between what we want to assess (the causal effect) and what we actually do assess using statistical methods. A mediator is the variable that explains the causal effect of X on Y (X>Z>Y). If you control for a mediator, you will conclude that there is no causal link, when there is.

There are several rules for controlling for possible confounders:

1. In a chain junction (A -> B -> C), controlling for B prevents information from A getting to C and vice versa.
2. In a fork or confounding junction (A <- B -> C), controlling for B prevents information from A getting to C and vice versa.
3. In a collider (A -> B <- C), controlling for B will allow information from A getting to C and vice versa.
4. Controlling for a mediator partially closes the stream of information. Controlling for a descendant of a collider partially opens the stream of information.

A variable that is associated with both X and Y is not necessarily a confounder.

A false positive refers to an incorrect rejection of the null hypothesis. The decisions a researcher can make during the research process is called the researcher degrees of freedom. Four common degrees of freedom are choosing sample size (1), using covariates (2), choosing among dependent variables (3) and reporting subsets of experimental conditions (4). The researcher degrees of freedom can significantly increase the false positive rate.

There are six guidelines for authors to prevent the increased rate of false positives:

1. Authors must decide the rule for terminating data collection before data collection begins and report this rule in the article.
2. Authors must collect at least 20 observations per cell or else provide a compelling cost-of-data collection justification.
3. Authors must list all variables collected in a study.
4. Authors must report all experimental conditions, including failed manipulations
5. If observations are eliminated, authors must also note what the statistical results are if those observations are included.
6. If an analysis includes a covariate, authors must report the statistical results of analysis without the covariate.

There are four guidelines for reviewers to prevent the increased rate of false positives:

1. Reviewers should ensure that authors follow the requirements.
2. Reviewers should be more tolerant of imperfection in results.
3. Reviewers should require authors to demonstrate that their results do not hinge on arbitrary analytic decisions.
4. If justifications of data collection or analysis are not compelling, reviewers should require the authors to conduct an exact replication.

A theory is a concise statement about how we believe the world to be. There are several things to look at when evaluating theories:

1. Descriptive adequacy

Does the theory accord with the available data?

2. Precision and interpretability

Is the theory described in a sufficiently precise fashion that it is easy to interpret?

3. Coherence and consistency

Are there logical flaws in the theory? Is it consistent with theories of other domains?

4. Prediction and falsifiability

Can the theory be falsified?

5. Postdiction and explanation

Does the theory provide a genuine explanation of existing results?

6. Parsimony

Is the theory as simple as possible?

7. Originality

Is the theory new or a restatement of an old theory?

8. Breadth

Does the theory apply to a broad range of phenomena?

9. Usability

Does the theory have applied implications?

10. Rationality

Are the claims of the theory reasonable?

Postdiction refers to predictions under controlled conditions. 

People do not reason logically, but decision making encompasses all human behaviour. Choices can be rational (internally consistent). People expect choices to show transitivity. If a relationship holds between item one and item two and between item two and item three then the relationship should also hold between item one and item three. Utility is the personal value we attach to outcomes rather than to their absolute monetary value.

Normative theories of decision making imply that some choices are better than other choices. The optimal choice in normative theories depends on the theory.

The expected value theory states that the optimal choice is the choice that offers the largest financial payoff. People also tend to go for choices with the maximum utility. People are not always consistent in their choices with expected value and utility, as this requires a lot of time and motivation. There are two principles in rational decision making:

1. Description invariance
   The description of the choice should not make any difference as long as basic structure of the choices is the same.
2. Procedure invariance
   The procedure of decision making should not make any difference in the decision that people make.

People are inconsistent with this. Psychic budgets refers to how we mentally categorize money that we have spent or are contemplating to spend. Sunk cost refers to an investment that is irretrievably spent and should not influence present decision making, but still does, as people want to get their investment out of it as much as they can. People also make decisions based on loss aversion, the unpleasantness of a loss is bigger than the pleasure of a similar gain. Therefore, people make decisions based on aversion of the unpleasantness of a loss. Satisficing refers to selecting the first choice that satisfies a certain demand (e.g: cost of a phone). People use satisficing to prevent them from having to compare everything with each other.

People tend to use heuristics to make decisions. There are several heuristics:

1. Representativeness heuristic
   An event is judged to be probable if it has properties that are representative of that category (e.g: we believe a person wearing a metal t-shirt is more likely to be part of a metal band than someone in a suit).
2. Availability heuristic
   An event is judged more probable if one is able to recall many examples of it (e.g: deadliness of plane crashes versus cardiovascular diseases).
3. Anchoring and adjustment heuristic
   The initial value of an event adjusts or estimate upwards or downwards on the basis of other information (e.g: if people first hear that someone wants an offer between 10 and 50 euros, they will have a lower offer than when people hear that someone wants an offer between 50 and 150 euros).

The odds of a conjunction of two events are always lower than the odds of a single event. The