Observation in clinical psychology refers to perception with the intention to draw conclusions. It can be used to gather information about others with whom a person communicates directly or indirectly (1), relationships and situations in which a person is either involved or not involved (2) and the person himself (3).

There is a distinction between every day observation and professional observation. Professional observation refers to goal-oriented observation on the basis of an observation question which leads to an informed decision to act. The method is first perception and then interpretation.

A rating scale refers to a written behavioural examination that draws on psychological knowledge (e.g. intake interview). An observation scale refers to a written behavioural examination that draws on observation. Subscales refer to components of the observation scale.  

Behavioural aspects can be accurately observed in a test situation. Observation during an interview can be useful for generating hypotheses about a client’s problems. The leniency effect (1), the halo effect (2), the logic error (3), the contrast error (4), the primacy effect (5), the recency effect (6) and the tendency to mostly give average scores and to avoid making extreme judgments (7).

The logic error refers to the tendency to pass similar judgements on traits that seem to be logically linked instead of an isolated impression. The psychologist’s theoretical background may also influence the fundamental attribution error.

Heuristics refer to information processing strategies that people use to reduce complex assessment tasks to simpler operations. There are two heuristics that influence observation in a clinical context:

1. Availability heuristic
   This refers to the tendency to give greater weight to more available or recent information.
2. First impression heuristic
   This refers to the tendency to hold on to one’s first impression even when there is additional information that contradicts the initial impression.

Standardized observation is generally selective (i.e. not everything is observed). The level at which the observation is observed needs to be determined. The molar level refers to larger, meaningful units. The molecular level refers to smaller behaviour units. The validity is high but the reliability is low at the molar level. The reliability is high but the validity is lower at the molecular level.

There are several potential settings for observation:

1. Natural surroundings
   This is observation in the setting in which behaviour normally occurs (e.g. school) The problems of observation in natural surroundings are the investment of time (1) and the information is not necessarily comparable (2).
2. Observation in simulated settings
   This is observation in a natural setting with a predefined task. The validity is high but the comparability with other clients is partially possible due to the semi-structured nature of the situation.
3. Laboratory observation
   This is observation in a setting which is artificial. The advantages are that there is a high degree of standardization (1) and the reliability and interrater reliability can be determined relatively easily (2). The disadvantages are the potentially low validity (1), it is very time-consuming (2) and can only be carried out with the use of certain equipment (3).

Event sampling is chosen when there is interest in the content or quality of a specific behaviour and possibly the frequency with which this behaviour occurs. This method requires the observation to be observed to be very clearly defined. Time sampling refers to observing behaviour at regular intervals regardless of which behaviour is being exhibited at that moment.

Self-observation refers to a method that is frequently used in behavioural therapy in order to gain an impression of the nature of the problem behaviour and the conditions under which it occurs. Self-observation is mainly suitable for measuring the frequency and nature of the occurrence of problem behaviour and is suitable for recording the frequency and nature of behaviour that are not visible to others. Self-observation is valid but not necessarily reliable.

Observation carried out by someone from the client’s immediate circle is comparable to self-observation in terms of advantages and disadvantages but an additional disadvantage is that the observation may be influenced by the affectionate relationship between the person and the client.

The behavioural avoidance test is used for the observation of avoidance behaviour. This method consists of compiling a list of situations with which the client has difficulties. The client will then arrange these situations according to difficulty. The advantage of the method is that the situations are based on the individual client but the disadvantage is that it is difficult to compare individual phobics.

Social anxiety consists of the extent to which a client avoids certain situations and the quality of his social skills. Midi-level measurements are measurements that attempt to combine the advantages of both methods (i.e. molar and molecular) in evaluation systems.

There are four categories of mental disorders:

1. Cognitive disorders (e.g. attention, memory)
2. Emotional problems and personality changes (e.g. anxiety, apathy)
3. (Senso)motor disorders and sensorimotor disorders (e.g. tremors)
4. Psychosocial disorders (e.g. loneliness; relationship problems).

The Nurses’ Observation Scale for Inpatient Evaluation (NOSIE) is capable of distinguishing more severely disturbed psychiatric clients and less severely disturbed psychiatric clients. However, it is unsatisfactory on reliability, validity and all other aspects.

When assessing the behaviour of older clients, there are a number of specific aspects that are significant (e.g. presence of symptoms of dementia).

Expressed emotion (EE) is a system that provides information about the degree of criticism and emotional over-involvement in a family member’s attitude towards a client. It is used for insight into the risk of psychotic relapse in a client with schizophrenia. It can be determined by using the Camberwell Family Interview (CFI) (1), using the Five Minute Speech Sample (FMSS) (2) or a combination of the both (3). Emotional overinvolvement is an important predictor for schizophrenia and can be an important predictor for psychopathology.