Terms in this set (111)
What is the most rigorous form of scientific inquiry?
The True Experiment
What is the purpose of a True Experiment?
To show cause and effect.
What three things must a True Experiment have?
Manipulation of Variables, Random Assignment, Control Group
What is an Confounding (extraneous) variable?
Any factor not directly related to the study but that may affect the dependent variable.
What are the two types of Confounding (extraneous) variables?
Intrinsic and Extrinsic
What is an Intrinsic Confounding Variable?
A factor that represents a personal characteristic of a subject in the study.
What is an Extrinsic Confounding Variable?
A factor that emerges outside the individual, from the environment or experimental situation.
In a study examining Anxiety there was a six week period between pre and post tests of anxiety levels. During this six weeks, the US invaded Iraq potentially causing more anxiety. What type of Confounding (extraneous) variable is this?
The control condition of an experiment is designed to control for what?
Confounding (extraneous) variables
What is an Active Variable in an experiment?
A variable that is manipulated, ie. subjects assigned to levels of the independent variable or treatments.
What is an Attribute Variable in an experiment?
A variable that is not manipulated but only observed. Ie. gender or age.
What is Random Assignment in an experiment?
Each subject has an equal chance of being assigned to any group or experimental condition.
Why are Random Assignment and Random Selection NOT the same thing?
Random selection is everyone in a population has an equal chance of being selected. So the sample represents the population as a whole.
Random assignment is each subject has an equal chance of being assigned to an experimental condition. The sample is NOT representative of the entire population.
T or F:
Differences between experimental groups are due to chance alone.
What do we mean by "purposefully decreasing intersubject differences?"
If we know that some of our participants are very good at the outcome variable already, we may split them apart and put them equally into each experimental group. One in the experimental and one in the control.
What are the five methods to control Inter-subject Differences in an experiment?
Subjects as their own Control
Analysis of Covariance (ANCOVA)
Controlling for gender and age in a study of cardiac disease, ie. using only older male participants, could be a possible breach of what ethical responsibility?
Principle of Justice
What is the major disadvantage to using a homogenous group of subjects for a study?
Results of the study can only be generalized to the type of subject that participated in the study.
ie. no cardiac research on women
How does Blocking work when controlling inter-subject differences?
If you use the extraneous variable of age as an independent variable. For example, experimental condition is further divided into three age categories, less than 30, 30-40, and over forty. This lets researchers also analyze the effect of age on the treatment condition too.
What is Matching? How does it control for inter-subject differences?
Matching is when all groups have equal numbers of subjects with specific characteristics. First, they arrangle people by characteristics of gender, age, etc. Then they randomly assign people from each group to different experimental conditions. This way all the groups have the same number of people with each characteristic.
What type of study has the highest level of matching?
Identical Twin Studies
What type of control for inter-subject differences occurs when all subjects receive all levels of the I.V.?
Using subjects as their own control
What do we call the I.V. during a repeated measures study?
Repeated Factor or Repeated Measure
What is one of the most efficient methods for controlling inter-subject differences?
What is an Analysis of Covariance (ANCOVA) ?
Method for controlling inter-subject differences that does not involve a design strategy but uses statistics to equate the groups on identified extraneous variables (covariates).
How does the ANCOVA work?
Removes the confounding effect of covariates by making them artificially equivalent across the groups, estimating what the D.V. would have been under equivalent conditions.
What is the key to using any method of controlling for inter-subject differences?
Correctly identifying the confounding variables.
What is the Control group?
Group of subjects that may receive no experimental condition, a placebo, or a standard treatment.
What must we be concerned with when a Control group gets no treatment or a placebo?
Ethical concerns. They must receive treatment the experiment is over.
What are the four types of Design Validity that relate to the four questions experiments seek to answer?
What are some possible threats to Statistical Validity?
Were appropriate statistics used?
Low statistical power = small sample size
Violation of statistical assumptions
Compounding error rates
ANOVA and ANCOVA are typically used when data is normally distributed. If these tests are used without data being normally distributed, it is a violation of a statistical assumption. What kind of design validity is threatened?
What do we mean by compounding error rates as a threat to statistical validity?
Repeated statistical tests on the same data can increase the probability of drawing incorrect conclusions.
What is Internal Validity?
Type of design validity that represents degree to which change in the D.V. is caused by change in the I.V.
What are some threats to Internal Validity?
History, Maturation, Attrition, Testing, Instrumentation, Regression toward the Mean, Interactions with Selection, Compensatory Equalization of Treatments, Compensatory Rivalry and Resentful Demoralization
How is History a threat to Internal Validity?
Refers to confounding effects of specific events in history occurring at the same time as research being conducted. ie. Desert Storm, 9/11, etc
How is Maturation a threat to Internal Validity?
Growth of subjects between first and second measurement. For example, patient grows older, stronger, healthier.
How is Attrition a threat to Internal Validity?
Number of people that drop out must be reported. Remaining group might no longer be representative of the sample.
How is Testing a threat to Internal Validity?
Patient's familiarity with the test is increased with every occurrence. Testing effects.
Why is Instrumentation a threat to Internal Validity?
Concerned with reliability of measurement. Observers could become more skilled between pre and post tests.
Why is Regression toward the Mean a threat to Internal Validity?
Extreme scores on the pretest may become closer to the mean on the post test. Are results do this phenomenon or actually due to D.V.?
Why are Interactions with Selection a threat to Internal Validity?
Because factors other than experimental intervention can influence post test differences. ie. if one group is all smokers, this could effect the outcome.
Why are Compensatory Equalization of Treatments a threat to Internal Validity?
Occurs when the experimental condition is considered desirable and researchers or those who work with control group will try to even out experiences by providing services to Control group. This minimizes differences between the groups.
What is Compensatory Rivalry and Resentful Demoralization and how does it threaten Internal Validity?
Subjects in the less desirable condition or the control strive work harder and achieve similar results as those in experimental (desirable) condition. This will decrease differences between the two groups.
Also possible when subjects in lesser desirable condition feel demoralized or resentful. This will increase differences between two groups.
How do we control for Compensatory Rivalry and Resentful Demoralization in a study?
Minimize contact between those in different experimental conditions / control.
What is Construct Validity and how does it relate to experimental design?
How can cause and effect relationship shown in data be applied to theoretical constructs? Generalizable? We must have data from multiple studies to explain cause and effect.
What is External Validity?
Degree to which results of study can be generalized beyond the exact conditions of the study.
Which type of Design Validity is more important when seeking cause and effect? Internal or External Validity?
In research design notation, what does the " O " mean?
Observation or Measurement
In research design notation, what does the " X " mean?
What does this mean?
O1 X O2
In research design notation, what does " R " mean?
In research design notation, what does " ----------- " mean?
Nonequivalent Control Group
What does this mean?
O1 X O2
---------------- O1 O2
Pretest and post test for experimental condition on top and control group on the bottom. Quasi experimental as no random assignment.
What does this mean?
R O1 X O2
Pretest and post test for experimental condition on top and control group on the bottom. Random assignment.
What is a Factorial Design?
Experimental design that incorporates two or more I.V.'s, with independent groups of subjects randomly assigned to various combinations of levels of the two variables.
How many I.V.'s does a Two-Way Factorial Design have?
How many I.V.'s does a Three-Way Factorial Design have?
In an experiment with two I.V.'s Location of exercise and Exercise Intensity, each I.V. has two levels. What kind of notation would you use to describe this Factorial Design?
2 x 2
What is a Randomized Block Design?
Experimental design used when an attribute variable, or blocking variable, is crosses with an active I.V.
Homogenous blocks of subjects are randomly assigned to levels of a manipulated treatment variable.
In an experiment,24 subjects were recruited including 12 men and 12 women. Each gender group was randomly assigned one of three doses of the drug being tested. So four men received 0.5, 4 men received 1.5, and 4 men received 3.0. (same for the women)
What kind of Randomized Block Design is this?
How many groups are there total?
2 x 3 Randomized Block Design
6 groups total
What is a Nested Design?
Research design when a third I.V. is "nested" inside another I.V.
For example, 10 therapists are examined in a research study and are further split up into two groups each with five therapists. One group is of more experienced therapists and the other group of less experienced therapists. So experience is "nested" within therapists.
What is a Repeated Measures Design?
Research Design where one group of subjects is tested under all conditions and each subject acts as their own control.
What is another name for a Repeated Measures Design?
What is a Latin Square?
A matrix composed of equal numbers of rows and columns used to designate random permutations for a Repeated Measures Design.
What is a Crossover Design?
Used when only two levels of an I.V. are repeated, controls for order effects. Half the subjects in Repeated Measures Design receive Treatment a followed by B, and half receive Treatment B followed by A.
What is a Sequential Clinical Trial?
Special approach to a randomized clinical trial that allows for continual analysis of data as it becomes available. Results accumulated as each subject is tested, experiment can be stopped at any point when evidence is strong enough to determine significant differences between treatments.
What is Single Subject Design?
Experiment with one subject that allows us to draw conclusions about effects of treatment based on responses of only that one patient under controlled conditions.
What are some advantages of Single Subject Design?
Appropriate for clinical populations
Suited for Heterogeneous populations
In a Single-Subject Design, what is the I.V.?
In a Single-Subject Design, what is the D.V. called?
Target Behavior that is observable, quantifiable, valid indicator of treatment effectiveness.
What are the two core elements of a Single Subject Design?
Repeated Measurement and Two Phases (Baseline and Intervention)
What are the two phases in a Single Subject Design? Describe them?
Baseline Phase - prior to treatment, control
Intervention Phase - after treatment
What is the baseline phase and what makes a strong one?
Baseline phase is prior to treatment and used as the control. It measures the natural state of the target behavior over time.
Must be monitored 3-5 times or until stability
What is a stable baseline?
Flat or moving in the direction opposite what is expected with the intervention.
If you saw an unstable baseline moving in the direction expected with intervention, what would say about the following graph?
I would say that the B phase is just a continuation of the A phase. It might look the same without the intervention.
What is a weakness of AB single subject designs?
Low internal validity. Did the I.V. really cause a change in the D.V.?
How can you increase the internal validity of an AB single subject design?
Repeating baseline and intervention phase. Using ABA or ABAB.
If the data in an ABA design changes in response to removal of the intervention during the second "A" phase, what kind of evidence does this offer?
Strength that the intervention caused the change. Increase internal validity.
What is a Therapeutic effect?
Effect of I.V. is maintained even after the intervention is withdrawn. Effect is not reversible.
What is the Prosthetic effect?
Effect of I.V. does not last after intervention is withdrawn. The effect is reversible.
What kind of design would have a baseline phase and two interventions?
What is a downside of Multiple Treatment Designs?
Allow comparison of adjacent phases only
What kind of single-subject design involves alternating between two interventions or treatment conditions?
Alternating Treatment Design
Describe a Multiple Baseline Across Subjects Design.
One intervention is applied to the same behavior for 3 or more subjects. Baseline is monitored for each until intervention introduces in staggered manner.
What is a concern of Multiple Baseline Across Subjects Designs?
Must control for outside influences as best you can.
What is an advantage of Multiple Baseline Across Subjects Designs?
Staggering intervention for each patient provides even more internal validity. Strong evidence that the intervention is working.
What is a Changing Criterion Design?
Begins with baseline but each subsequent treatment phase has a higher criterion for achievement. Previous intervention phase serves as baseline for next.
How is reliability improved in single subject designs?
Having two testers simultaneously observe target behavior and calculating agreement score between them.
When evaluating data with Visual Inspection, what are you looking for?
Stability and Direction within a phase.
Between phases we look for level (value of DV), trend (direction of change), and slope (rate of change).
How do we calculate a celeration line for single subject design?
Separate for baseline and intervention.
Divide number of data points in half on x axis, separate with vertical line. Divide two halves in half again and determine median score to divide data of each in half on y axis. Draw line connecting the two points of intersection (horizontal with verticle).
What is the Split-middle technique?
When a celeration line is developed in Phase A and extended into Phase B.
If the intervention had no effect, how many data points would you expect above and below the acceleration line in the intervention phase?
50 / 50 above and below
What should we remember about Split Middle technique and evaluating Single Subject Design data?
Difference between data points around celeration lines is due to treatment and not to chance.
What is Descriptive Research?
Designed to document factors that describe characteristics, behaviors and conditions of individuals or groups.
What is Qualitative Research?
Research that seeks to describe the complex nature of humans and how individuals perceive their own experiences within a specific social context.
What is the purpose of Qualitative Research?
Describe and explain complex human experiences from the perspective of the person. Deep understanding of particular.
What is Idiographic Data?
Data that relates to a given case in a given time.
What are the two types of Qualitative Research?
Phenomenology and Ethnography
What is Phenomenology?
Seeks to explain complex reality of particular phenomenon through analysis of people who experience it.
What is Ethnography?
Seeks to explain social world of a particular group, researcher becomes immersed in the social group's way of life.
Give an example of Ethnography.
Margaret Mead and "Black Like Me"
How do you select subjects in Qualitative Research?
By selecting those who have the best chance of providing insight.
What is observation in Qualitative Research?
Participant observation - researcher becomes apart of group being studied.
How are Interviews used in Qualitative Research?
3 ways: Structured, Semi-structured, Unstructured
What is an Artificat in Qualitative Research?
A specific object relating to the phenomenon being studied.
What is Triangulation?
Process by which concepts are confirmed by using more than one source of data, more than one data collection method, more than one group of researchers.
Triangulation uses multiple sources of information. How is it analyzed?
Multiple researchers independently code the data.
What is reliability in Qualitative Research?
What is Validity in Qualitative Research?
What is External Validity in Qualitative Research?
YOU MIGHT ALSO LIKE...
Experimental Research and Single Subject Designs
HCOM 500 CH. 4 Research Design
OTHER SETS BY THIS CREATOR
Comps - Functional Training, Equipment, and Devices 1 (GAIT)
Comps - Cardiovascular 2
Comps - Cardiovascular 1
Comps - Musculoskeletal 2
THIS SET IS OFTEN IN FOLDERS WITH...
Physical Agents - Inflammation and Thermodynamics
Physical Agents - Heat and Ultrasound