Bias

What is compensatory rivalry and resentful demoralization?

CR - "we'll show them" attitude RD - "we're getting the short end of the stick, why bother" attitude due to communication between groups

What is interpretation bias?

allowing interpretation to foster the findings in a way the data does not support

What is the Hawthorne effect?

artificial change produced by the measurement process individuals perform better when they know they are in a study regardless of intervention

What is publication bias?

bias associated with reporting that involves publishing findings that are difference-oriented (vs a finding of no significance) and the delay to publication

What is temporal precedence?

cause must precede the effect if effect occurs before cause -> can it be the cause

What are examples of threats related to investigators?

compensatory equalization of treatments experimenter bias

What are types of errors in clinical judgment?

confirmation bias recency effect representative exclusivity value bias recall bias

What is the history effect?

confounding events occurring in between pre- and post-testing subject activity changes; global events; environment changes

What is contamination bias?

diffusion or imitation of treatment members of one group receive treatment or are exposed to intervention meant for the other group

What are ways to solve for regression to a common mean?

eliminate outliers from baseline scores take repeated baseline scores and average them together

What are ways to solve for compliance bias?

explain importance of adhering to routine clear and concise instructions

What is testing bias?

familiarity with, inaccurate or biased collection of measures, during the data gathering process different instructions/cues lack of familiarity with testing procedure

What is internal validity?

focuses on cause and effect relationship represents the extent to which we can accurately state that the independent variable produced the observed effect

What are interaction effects?

generalizability is related to specific patient context and conditions under investigation interactions are effects caused by combinations of treatment and other investigation conditions

What are ways to solve for testing bias?

give subjects several practice sessions average scores of several trials from one session protocol for administering test ensure competency of administrators prior to collecting data

What are examples of threats related to study logistics?

history effect instrumentation bias testing bias

What is model validity?

how representative is the treatment "model" compared to clinical practice

What are ways to solve the Hawthorne effect?

importance of truthful responses and performance repeated measures and average them together

How do you control for bias?

inclusion/exclusion criteria allocation design types of outcomes and outcome assessment enrollment/treatment/measuring/running stats reporting standards checklist

What is compensatory equalization of treatments?

individuals providing treatment in study purposefully or inadvertently supplement activity in control group to "make up" for what the intervention group is receiving issue in studies with multiple providers

What are ways to solve compensatory rivalry and resentful demoralization?

keep subjects from each group separate mask/blind subjects explain importance of adhering to routine

What are ways to solve contamination bias?

keep subjects from each group separate mask/blind subjects explain importance of adhering to routine

What are ways to solve selection/sampling bias?

larger population size randomized sampling implementing research design methods to overcome bias collect data at multiple sites

What are ways to solve compensatory equalization of treatments?

mask investigators/therapists so they do not know which group is receiving which intervention provide clear and explicit protocol for administration limitation of communication about interventions between providers or treatment

How to resolve covariation of cause and effect?

must be able to show that effect only occurs in presence of cause degree of outcome is related to magnitude of intervention

What is regression to a common mean?

natural progression of conditions poor instrument reliability

What is maturation bias?

occurs due to the passage of time (between pre-/post-test) physical, psychological, emotional, spiritual unrelated to therapeutic intervention

What is compliance bias?

occurs when differences in subject adherence to the planned treatment regime or intervention affects the study's outcomes

What is reporting bias?

occurs when researchers choose to not report on all outcomes measures because of no or negative effects

What is attrition bias?

occurs when subjects who leave the study differ from those that remain affects generalizability to population

What is selection/sampling bias?

occurs when the exposure and no exposure groups have different risks of developing the outcome of interest for reasons other than being exposed measure of effect of exposure is distorted because of the association of the study factor with other factors that influence the outcome

What are ways to solve instrumentation bias?

proper usage/understanding/calibration protocols for data collection maintain constant conditions orientation/training on usage

What are ways to solve for experimenter bias?

protocol for coaching and cueing during administration blind therapists from intervention being given (if possible)

What are other forms of bias

publication bias interpretation bias reporting bias

What are ways to solve allocation/assignment bias?

randomized allocation techniques statistical adjustment for baseline values if randomization is not possible

What is external validity?

relates to generalizability of findings can findings be generalized to situations different from experiment

How to resolve plausible alternative explanations?

remove effect of confounding variables

What are ways to solve attrition bias?

replacement of lost subjects document characteristics and reason for leaving reexamination or remaining subjects estimation of removed statistics

What is experimenter bias?

researchers influence outcomes more positive interaction with experimental group observer's beliefs about therapeutic effectiveness may influence outcomes and their measurements

What is bias?

results or inference that systematically deviate from the truth or the process leading to such deviation result of error in design or conduct of a study related to internal and external validity (reduce confidence in cause-effect and generalizability of results)

What are examples of threats related to subjects?

sampling/selection allocation/assignment attrition maturation compensatory rivalry/resentful demoralization contamination regression toward a common mean compliance Hawthorne effect

What is instrumentation bias?

selecting wrong tool limitations of tool malfunction of tool inaccurate application

What are ways to solve for maturation bias?

take baseline measures randomized allocation consistency of timing for data collection

What are the components of causality assumption?

temporal precedence covariation of cause and effect no plausible assumptions of confounding variables

How is independent validity achevied?

the controls placed within the study reduced confounding the true effect effect on dependent variable is only due to independent variable manipulation

What are the types of bias?

threats to subjects threats to investigators threats to study logistics

What is allocation/assignment bias?

typically a result from non-randomization how participants were assigned to groups in the study group characteristics may be different (external factors)

What are ways to solve the history effect?

use a control group and randomly assign subjects schedule the study to avoid a predictable external event