Quiz on Chapters 9, 10, 11 Flashcards
What are the characteristics of true experiments
- Independent variable must be manipulated by the examiner
- Subjects randomly assigned to groups
- Must have a control group
Describe the process and purpose of random assignment
Process: Use a table of random numbers, Subjects are numbered 01-45, use pairs of digits to identify them from the table, as they are chosen they are assigned to group 1, 2, or 3 on a rotating basis until all have been assigned
Purpose: To ensure that a sample is representative of the parent population and that it is not biased
Describe Various reasons for missing data
- Subjects may drop out or terminate treatment before the study is complete
- Subjects may cross over to another treatment during the course of the study
- Subjects may refuse assigned treatment after allocation
- Subjects may not be compliant with treatment
- Subjects may be excluded after randomization because they don’t meet eligibility criteria
Explain the rationale for intention to treat analysis (ITT)
- Guards against the potential for bias if dropouts are related to outcomes or group assignment and preserves the original balance of random assignment
- This approach is considered reflective of routine clinical situations in which some patients will not be compliant
Describe the purpose of blinding
- Protection against observation bias of patients, treating providers, measuring providers, and data analysts
- Observation bias: Participants knowledge of their treatment or the investigators expectations can influence performance and/or recording and reporting of outcomes
Describe strategies for controlling inter-subject differences in research design
- Selection of homogeneous subjects
- Only Males - Blocking
- Block of Males, Block of Females - Matching
- Matched for age and sex - Using subjects as their own control
- Analysis of Covariance (ANCOVA)
Describe the elements of statistical conclusion validity
- Is there is a relationship between the independent and dependent variables?
- Was there the potential that inappropriate statistical procedures were used for analyzing data, and did this lead to invalid conclusions about that relationship?
Describe the elements of internal validity
- Is there evidence of a causal relationship between independent and dependent variables?
- Refers to the potential for confounding factors to interfere with that relationship
Describe the elements of Construct validity
- To what theoretical constructs can the results be generalized?
- Refers to the theoretical conceptualization of the independent and dependent variables
Describe the elements of External Validity
- Can the results be generalized to other persons, settings, or times?
- Refers to the extent to which results of a study can be generalized outside the experimental situation
Discuss how experimental designs control for threats to internal validity
- Randomization is primary way
- If only 2 levels, counterbalancing is used
- If 3 or more levels, use a latin square
What is the difference between efficacy and effectiveness
Efficacy: The benefit of an intervention as compared to a control or standard program
Effectiveness: The benefits and use of the procedure under “real world” conditions
What is the difference between quasi experimental designs and true experimental designs
Quasi Experimental designs are similar to experimental but they lack randomization, comparison group, or both
Judge the extent to which threats to internal validity limit the application of quasi-experimental designs
While threats to internal validity exist in quasi experimental designs, the rigid structure and rules of the RCT do not represent real world situations, making it difficult for clinicians to apply or generalize research findings. The results of the RCT may not apply to a particular patient who does not meet inclusion or exclusion criteria, or who cannot be randomly assigned to a treatment protocol. Many of the quasi experimental models will provide an opportunity to look at comparison in a more natural context.
Describe a pretest - posttest control group design (independent)
- Used to compare two or more groups that are formed by random assignment
- One group receives the experimental variable and the other acts as a control
Describe a two group pretest - post-test design (independent)
- Incorporates two experimental groups formed by random assignment
- Used when a true control condition is not feasible or ethical, often comparing a “new” treatment with an “old” standard or alternative treatment
Describe a multigroup pretest- post-test control group design (independent)
- Similar to pretest posttest control group but with multiple experimental groups and a control group
- allows researchers to compare several treatment and control conditions
Describe a post-test only control group design (independent)
- Experimental group and a control group with post-test only
- Identical to pretest post-test control group with no pre-test
What is the difference between an independent and repeated measures group design
Independent designs have more than one group whereas repeated measures designs involve one group of subjects being tested under all conditions and each subject acts as their own control
Describe factorial design (independent)
- Incorporates two or more independent variables
- Described according to dimensions or number or factors
- Can also be described by number of levels within a factor
- 3 x 3 design includes two variables each with three levels
- 2x3x4 design includes three variables with 2, 3, and 4 levels respectively
Describe a two way factorial design (independent)
- Involves two independent variables
- ie: 2x3 design, two independent variables with 2 and 3 levels respectively
Describe a randomized block design (independent)
- When extraneous factor may influence differences between groups, can control for this by building the extraneous variable into the design as an independent variable
- Often used with an attribute variable as the blocking variable
- Homogeneous blocks of subjects are randomly assigned to levels of a manipulated treatment variable
Describe a single factor (one way) repeated measures design
- a single factor experiment where one group is exposed to all levels of one independent variable
Describe a cross over design
- To control for order effects, counterbalance treatment conditions so that the order is systematically varied
- Half the subjects receive treatment A followed by B; The other other receive B followed by A
Describe a (multi factor) two way design with two repeated measures
- Two factor experiment where each subject is exposed to all levels
- 2x2 design with one group, all exposed to each treatment
Describe a mixed model design
- Two independent variables, one repeated across all subjects, the other randomized to independent groups
What is a sequential clinical trial
- Special approach to RCT
- Continuous analysis of data and trial ends when results are conclusive and terminal decision is made to recommend A or B
Describe a one group pretest post-test design
- Quasi Experimental design that involves one set of repeated measurements taken before and after treatment on one group of subjects
- No comparison group makes this design weak, as well as history and maturation effects, decrease internal validity
Describe a one way repeated measures design over time
- Quasi experimental design, Multiple measurements of the dependent variable are taken within prescribed time intervals. The intervention may be applied once or may be repeated in between measurements
- Time is the independent variable and every subject is evaluated at each time interval, making it a repeated measure
- Internal validity is threatened due to lack of control group
Describe an Interrupted Time Series design
- Quasi experimental design, based on the application of multiple measurements before and after treatment to document patterns or trends of behavior
- involves a series of measurements over time that are interrupted by one or more treatment occasions
- Independent Variable is time
How can an Interrupted Tim Series design offer more control than a one group pretest post-test design
- The multiple pretests and post-tests act as a pseudo-control condition, demonstrating maturational trends that naturally occur in the data or the confounding effects of extraneous variables.
Describe a Non Equivalent Pretest Post-test control group design
- Quasi Experimental; Similar to the (independent) pretest post-test design but lacks randomization between groups
- Includes a pretest and a control group but no randomization
- Groups may be selected based on intact groups or because of subject preferences
- Major threat to internal validity is the interaction of selection with history and maturation
Describe a Non Equivalent design with historical controls
- Quasi Experimental Design; Uses historical controls who received a different treatment during an earlier time period
- May be used when ethical concerns preclude the use of a true control group
- There is potential for confounding because of imbalances in characteristics of the experimental and historical control groups
Describe a Non Equivalent Post-test only control group designs
- Quasi experimental design; identical to non equivalent pre-test post- test control group design, but with no pre-test
- An experimental group and control group are identified (not random) receive intervention or placebo and the post-test is performed
- Internal validity is threatened by selection bias and attrition
