Nomothetic
Pertaining to or involving the study or formulation of general scientific laws.
'nomothete': lawgiver
Idiographic
Pertaining to or involving the study or explication of individual cases or events.
'idios': own, or private
Three types of investigation for the clinical researcher
Case Study: typically focuses on one person/patient.
Correlational method:
Experimental method:
Case study pros and cons
+ Allow observation of a rare or infrequent situation
+ Can generate new ideas, perspectives and techniques
+ Challenge existing assumptions
- Observers are typically biased
- Situations and outcomes can be subjective
- Inherently idiographic
Correlational method proc and cons
+ External validity (large samples)
+ Results can be replicated
- No explanation of the correlation, only observation of its occurrence
Correlational method
The attempt to demonstrate that changes in specific variables have a direct influence on other specific variables.
Typically (on order for the study to be valid) a sample must be representative of the group the researcher intends to understand. (A study o
Internal Validity
The extent to which the results of a study can be attributed to a specific and defined course of action or sequence of events. (A caused B, or C was rectified by D.)
External Validity
When findings can be generalized outside the study.
Positive Correlation
When both variables (ie, quantitative fear of dogs and number of times bit by dogs) increase together.
Negative Correlation
When one variable increases while the other decreases. It should be noted that this is often just a factor of semantics, as the wording can completely change the variable measured. Fear of dogs increases as number of bites does. Trust of dogs DECREASES as
Magnitude
The degree to which two variables change in relation to one another. The greater the magnitude, the more closely a change in one variable indicates a change in the other.
Correlation Coefficient
A quantitative tool in relaying data, expressed as 'r'
+1.00 indicates a perfect positive correlation (1:1)
0.0 indicates no relation (0:1)
-1.00 indicates a perfect negative correlation (1:1)
Statistical Significance
A way of quantifying research data that indicates a strong probability that the results are not due to chance.
Typically, an analysis must indicate there is less than a 5% probability that results are due to chance in order for a study to be considered co
Epidemiological Study
A correlational study used to describe the incidence and prevalence of an occurrence within a group.
(How often do dentists commit suicide?)
Incidence
Number of novel occurrences during a given period.
Prevalence
Total number of cases within a group during a given period.
Longitudinal Study
A correlational study in which the same subjects are observed on numerous occasions over a long period of time. (Years or even decades)
Experimental Method
A type of research in which one variable is manipulated (independent variable) in order to observe a potential causal relationship with another variable (dependent variable). ie; does doing A make B happen?
Confound
A variable other than the 'independent' which also has a causal relationship with the 'dependent'. These must be eliminated to ensure the experiment shows only the relationship between independent and dependent variables.
Confound Elimination Tools
Control Group
Random Assignment
Blind Design
Control Group
In an experiment, the group upon which no action is taken. Or, the group not exposed to the independent variable. This group is used as a comparison (or baseline) for the experimental group.
Experimental Group
In an experiment, the group exposed to the independent variable. Any change in this group is compared to the control group to determine that the independent variable is the cause.
Random Assignment
In order to eliminate the possibility that a commonality amongst subjects might influence the dependent variable, subjects are randomly assigned to the groups.
Blind Design
In an experiment, intentionally preventing subjects from knowing which group (control, experimental) they are assigned. This is done in an effort to eliminate any bias they may have from influencing the outcome.
Double-Blind Design
In an experiment, intentionally preventing both subjects and experimenters from knowing to which group (control, experimental) the subjects are assigned. This is done in an effort to eliminate any bias they may have from influencing the outcome.
'Quasi Experimental' or 'Mixed' Design
When the experiment does not make use of random assignment.
This is necessary when ethical considerations prevent broader samples (to study the effect of therapy on rape victims, experimenters must use existing victims...we all hope) Also when it is impos
Natural Experiment
A type of quasi experiment in which the groups studied have been selected by natural events. Flood victims, radiation exposure, earthquake effects. Can't be willfully repeated due to unpredictable selection.
Analogue Experiment
When subjects are induced to replicate target behaviors, symptoms or conditions in order to conduct the experiment. Animals used in place of humans are also an example of this.
Inherently flawed, as it is impossible to be certain that the replicated condi
Single-Subject Experiment
An experiment in which the subjects condition prior to the introduction of the independent variable is used as the baseline.
Often used when a condition is extremely rare.
ABAB Design
A type of single-subject experiment in which the independent variable is removed after introduction, then introduced a second time. All four states are measured (pre-variable, variable present, variable removed, variable returned) in order to ensure that
Multiple Baseline Design
A single-subject experiment in which an independent variable is used to manipulate more than one dependent variable. Typically one dependent variable is focused upon at a time.
Example: using a reward to eliminate an unwanted behavior, then using the same
Clinical Significance
A measure of whether the amount of improvement (due to treatment) is meaningful to the individual's life. Typically can only be determined by the patient and therapist.