Independent variable: experimental manipulation/variable we are
evaluating to see if it has an effect
Dependent variable: outcome/measure we are examining to
reflect the impact/effects of independent variable
2.3 Treats to internal validity
- History
Any event (other than independent variable) during the experiment and common to all subjects which may
account for the results (natural disaster, fire drill etc.)
o Systematic bias/experience that subjects receive that could explain how they responded on
dependent measures
o Only threat to internal validity if it impacts the results
▪ If both groups are exposed to the event, the event itself can’t explain group differences
- Maturation
Processes within the participants that change over time (growing older, stronger, healthier, smarter)
o Only a problem if the design can’t separate effects of maturational changes from the intervention
o Maturation has a greater impact over time
o Multiple baseline measures control for maturation
- Testing
Effects that taking a test one time may have on subsequent performance on the test
o Practice/familiarity with test or measures influence performance on second testing
o Repeated testing doesn’t always lead to improvement
- Instrumentation
Changes in measuring instrument or measurement procedures over time
o E.g., changes in scoring criteria, definitions and instruments
o Changes in definitions are part of instrumentation when rates of assessment are used and compared
over time
o Contextual influences can also greatly influence outcome measures
- Response shift
Changes in a person’s internal standards of measurement/change in threshold for answering in a particular
way/changes in measurement device or how measure is used
o Threshold may be influenced by historical and maturational changes in the individual or the context
, - Statistical Regression
Tendency for extreme scores on any measures to revert toward the mean of a distribution when the
measurement device is re-administered (scores tend to be less extreme on re-testing)
o How to protect against statistical regression
▪ Random assignment of participants to conditions (regression affects all groups)
▪ Use measures with high reliability and validity (regression is a function of error of the
measure. The greater the error, the more likely there will be regression)
▪ Test everyone twice for the pretest and select only those individuals who were extreme on
both occasions. (feasible, so not used often)
▪ Multiple baseline design
- Selection biases
Systematic differences between groups before any experimental manipulation or intervention is presented
o Based on assignment of subjects to groups, they are already different from each other in important
way that might contribute to or explain results
o Random assignment of subjects is often used to minimize the likelihood of selection biases
- Attrition
Loss of subjects over the course of an experiment, can change the composition of groups
o Changes in overall group performance may be due to loss of subjects instead of the intervention
o Keep the study short and provide incentives for completing measures
- Diffusion or imitation of treatment
If the intervention is accidently provided to some subjects in the control group, or if a subject in the
intervention condition doesn’t receive treatment
o Leads to smaller (or not even detectable) differences between groups
- Special treatment or reactions of control
Special circumstance in which an intervention program is evaluated and provided to an experimental group,
but the no-intervention control group receives some special attention that can contribute to the results
o This can be the case if people find out that they are in the control condition and therefore no longer
want to participate in the study
o If participants find out that they are in the control group, they may react in ways that obscure the
differences between treatment and no treatment
2.6 Circumstances in which threats to internal validity emerge
- Poorly designed study
One in which from the outset we know that many threats will be plausible
o Pilot studies help to fine out the details of a study and develop a controlled study in which
inferences about intervention effect can be drawn
- Well-designed study but sloppily conducted
Random assignment of participants, but in the conduction there is diffusion of treatment
- Well-designed study with influences hard to control during the study
If circumstances occur that are not easily controlled by the investigator (attrition)
- Well-designed study but results obscure drawing conclusions
Between-groups differences did not meet statistical significance. Results make any conclusions ambiguous.
o History, maturation, testing and statistical regression can explain the results
2.8 External validity
Plausible factor that restricts the generality of the results (of the intended! Population)
Threats to external validity constitute questions that can be raised about the limits of the findings.
- Does this apply to something that is normally encountered and not part of laboratory setup
- Does it apply to other groups of persons or other geographical areas
- What are the boundaries or limits of the demonstrated relationship
Major threats to external validity
- Sample characteristics
To what extent can results be generalized to others who vary in characteristics (race, age, background)
o Translating findings of animal studies to human studies
o College students as subjects/heavy reliance on undergraduates as subjects
, ▪ College students are WEIRD: Western, Educated, Industrialized, Rich and from Democratic
Cultures. They don’t represent other populations in fundamental ways such as attributions,
reasoning style, personality, perception
o Samples of convenience: Selection of subjects merely because they are available
o Underrepresented groups (women, ethnic minorities)
- Narrow stimulus sampling
If the findings occurred under the very restricted/narrow stimulus conditions of the study
o One experimenter/therapist, one story/stimulus
- Reactivity of experimental arrangements
Being aware of participating and responding differently as a result
o Would the results be obtained if subjects were not aware that they were being studied?
- Reactivity of assessment
Extent to which subjects are aware that behaviour is being assessed and this influences their responses
o Obtrusive: ff subjects are aware that performance is being assessed
o Reactive: if awareness of assessment leads persons to respond differently
▪ Experimental arrangement: whether subjects believe they are participating in experiment
or (unconsciously) pick up cues that guide their behavior
▪ Assessment procedures: whether subjects are aware of being assessed and presumably can
alter their performance as a result
o You can combat this with including a measure where purposes are not so clear and therefore
distortion is less likely
- Test sensitization
Administration of pretest may sensitize subjects so that they are affected differently by the intervention.
o Results may not generalize to subjects who did not receive the pretest
- Multiple treatment interference
Drawing conclusions about a given manipulation when it’s evaluated in the context of other manipulations
o When more than one task is presented, performance of second task can be influenced by prior task
o Can be solve by balancing the order of conditions among participants
- Novelty effects
Effects of the intervention may depend upon and be limited to the context in which it is administered
o Context may make experimental manipulation salient or novel in some way
- Generality across measures, setting and time
o were results obtained on measures unique to particular measure instead other factors of the
construct
o To what extent are findings likely to be restricted to that setting
Cohort: particular group of people who have shared something over a particular time period
Proof of concept/test principle: A test to see whether something could occur that is important in principle or in
theory
Voordelen van het kopen van samenvattingen bij Stuvia op een rij:
Verzekerd van kwaliteit door reviews
Stuvia-klanten hebben meer dan 700.000 samenvattingen beoordeeld. Zo weet je zeker dat je de beste documenten koopt!
Snel en makkelijk kopen
Je betaalt supersnel en eenmalig met iDeal, creditcard of Stuvia-tegoed voor de samenvatting. Zonder lidmaatschap.
Focus op de essentie
Samenvattingen worden geschreven voor en door anderen. Daarom zijn de samenvattingen altijd betrouwbaar en actueel. Zo kom je snel tot de kern!
Veelgestelde vragen
Wat krijg ik als ik dit document koop?
Je krijgt een PDF, die direct beschikbaar is na je aankoop. Het gekochte document is altijd, overal en oneindig toegankelijk via je profiel.
Tevredenheidsgarantie: hoe werkt dat?
Onze tevredenheidsgarantie zorgt ervoor dat je altijd een studiedocument vindt dat goed bij je past. Je vult een formulier in en onze klantenservice regelt de rest.
Van wie koop ik deze samenvatting?
Stuvia is een marktplaats, je koop dit document dus niet van ons, maar van verkoper manonschoonderwaldt. Stuvia faciliteert de betaling aan de verkoper.
Zit ik meteen vast aan een abonnement?
Nee, je koopt alleen deze samenvatting voor €7,49. Je zit daarna nergens aan vast.