Lit Eval Exam 2

What is Chi-square often used to evaluate?

similarity between treatment and control groups

What type of data is used for chi square and what is an example?

frequency data; number of people

What types of variables are used chi-square testing?

nominal, ordinal

What is a t-test used to bind differences between?

2 groups on a ratio/interval variable

What does a paired t-test look at?

2 groups before and after

What are the two variables in paired t-tests?

one is time-related and second is measured at each of those times

What is variable that is time-related for a paired t-test?

pre-intervention group vs. post-intervention group

What type of data is the second variable for a paired test?

interval, ratio

What does doing multiple tests increase the risk for?

type I error

What is an analysis of variance test and how is it different than a t-test?

provides same information as t-test except more than 2 groups compared

Why would you rather do an ANOVA test instead of multiple t-tests?

alpha increases

What type of data is the variable of interest for ANOVA?

interval/ratio

What does n stand for in the N-way ANOVA?

number of independent variables

For ANOVA what does the dependent variable equal?

independent variable and error

What do the interval/ratio level data equal for ANOVA?

nominal or ordinal and error

If an ANOVA test comes back with a significant difference, what is the next step?

determine which groups differ from each other

What is the type of test done after an ANOVA test to determine if groups are different?

multiple comparison test

What does ANOVA allowed modeling of?

complex issues

What does R^2 tell you for an ANOVA test?

how good the model is

What's in the error term for an ANOVA test?

variables you left out of your model

What are other stats that are similar to ANOVA?

regression, MANOVA

What are the independent and dependent variables for a regression test?

interval/ratio

What is a MANOVA test?

same as ANOVA but with >1 dependent variable

What is a good way to think of ANOVA?

expanded t-test test, allows for modeling several variables

What does the way research is designed influence?

quality

What type of research design is a single group observed after exposure, which is presumed to cause a change?

one-shot case study

What type of research design would a RPh giving an in-service on Lyme disease to other RPhs and then testing their knowledge 1 week later be?

one-shot case study/case report

Why does a case study not have group wise?

no comparison group

What are threats to a case study?

history, maturation, selection bias, mortality

What kind of research design is it when a single group is pretested, exposed to treatment, and post-tested and then the difference between pre and post are attributed to the intervention?

one-group pretest-posttest design

What type of research design is cholesterol measured, started on a statin, and measured 6 months later?

one-group pretest-posttest design

Does a one-group pretest-posttest design have a control group?

no comparison group

What are the threats of validity to one-group pretest-posttest design?

history, maturation, testing, instrumentation, regression to mean

What type of research design is it when a comparison is made between two groups and only 1 gets treatment?

static group design

What type of research design is 50 volunteers taking a statin and then measured. Later, a second group have their chol. measured?

static group design

What are the threats to validity for a static group design?

selection bias, mortality

What are the problems with pseudo-experimental designs?

uncontrolled extraneous variables, threaten validity, make cause and effect illusive

What do true experimental designs have built in and what does this mean?

controls for validity threats; relatively sure the indep. variable was cause of difference

What type of research design is it when 2 groups measure the outcome, they are measured twice, and they are randomly assigned to each group?

pretest-posttest control group design

When assigning people into random groups, what are we assuming?

assumed to be equivalent at the start of the study

When are measurements taken for a pretest-posttest control group design?

at the same time

What does one group receive for a pretest-posttest control group design and what does the other receive?

treatment; placebo/standard care

What controls for selection bias in research designs?

randomization

What controls for history and maturation by having it affect each group equally?

equal groups

What controls for regression to the mean in research design?

groups selected equally

What controls for the testing effect by having it be equal for both groups?

groups selected equally

What controls for instrumentation problems in research design?

same procedure and time

What is one threat to validity thats impossible to control for?

mortality

What if the pretest sensitizes subjects to the treatment?

solomon four group design

In Solomon Four Group design, who is pretested?

one experimental group, one control group

How many control groups are there for Solomon Four Group design?

2

How many groups are post-tested for Solomon Four Group design?

all 4

What does Solomon Four Group Design control for, what doesn't it control for, and what does it require?

all except mortality; requires more subjects

What did Willowbrook lead to for informed consent?

Civil Rights for Institutionalized persons act of 1980

What did the Tuskegee Syphilis study lead to for informed consent?

HHS policy for protection of human subjects

What did the Nuremberg trials lead to for informed consent?

Nuremberg code - 10 directives or human research

Based on the Nuremberg Code what is absolutely essential for human subjects?

voluntary consent

According to the Nuremberg Code, what must an experiment yield?

generalizable knowledge that could not be obtained in any other way and is not random and unnecessary in nature

What should precede human experimentation based on the Nuremberg code?

animal experimentation

What should be avoided based on the Nuremberg code?

all unnecessary physical and mental suffering and injury

When should no experiment be conducted based on the Nuremberg code?

reason to believe that death or disabling injury will occur

What should the degree of risk to the subjects never exceed for the Nuremburg Code?

humanitarian importance of problem

What should risk be based on the Nuremberg code?

minimized

Who should only conduct experiments based on the Nurenburg Codes?

scientifically qualified investigators

What should subjects always be at liberty to do based on the Nurenburg code?

withdraw from experiments

Investigators must be ready to end the experiment at any stage if what occurs based on the Nuremburg laws?

if cause to believe that continuing the experiment is likely to result in injury, disability, or death to the subject

What are the 3 underlying ethical principles that guide ethical conduct of research involving humans?

respect for persons, beneficence, justice

What ethical principle is it when subjects are capable of deliberation and researchers respect their opinions and choices and don't obstruct their decisions and choices?

respect of persons

Which ethical principle is it when researchers make an effort to secure the well being of subjects, maximize possible benefits and minimize possible harm, and strive for societal benefits but not at the expense of individual harm?

beneficence

What ethical principle is it when participants are treated fairly, don't include people who are unlikely to benefit from research, and don't choose subjects simply because of their availability, vulnerability, or compromised position?

justice

Who are in clinical trials, what are they designed to assess, what do they often include, and what do they require?

human subjects; assess safety/efficacy, include controls, IRB approval

Which phase of clinical trials is determining safe dose range, tolerance, kinetic parameters in humans?

phase I

Who is normally included in phase I clinical trials?

normal, healthy males ages 20-45

How many participants are involved in Phase I clinical trials?

20-80

What phase of clinical trails is it when they are determining safety and efficacy in small number of patients?

phase II

What phase of clinical trials do they start to have specific inclusion/exclusion criteria?

phase II

How many participants are involved in phase II clinical trials?

200

Which phase of clinical trials is it when determining additional efficacy information for a specific disease and on ADRs?

phase III

What may patients have in phase III clinical trials?

concomitant diseases

How many participants are in phase III clinical trials?

100-1000s

What phase of clinical trials is it when post-FDA approval studies for safety and additional indications occur?

Phase IV

What else might they look at in Phase IV clinical trials?

long term safety, drug interactions

What is informed consent?

explicit rules and regulations to ensure protection of patients' rights

What do the IRBs do and who are they required by?

determine potential risks and benefits by reviewing the protocol; FDA

What is used by FDA as basis for informed consent?

Declaration of Helsinki

What is the probability of finding a difference when there really is a difference?

power

How do you calculate power level?

1-B

What does Beta equal in the power equation?

type II error

What 4 factors do power depend on?

sample size, chosen alpha level, effect size, reliability of the measurement

The larger the sample, the _____ powerful the test?

more

What happens to Beta as sample size goes up?

down

What can you do to sample size if you increase alpha?

decrease

What is the trade-off of decreasing sample size by increasing alpha?

more chance of type I error

What is the effect size?

change you are interested in

If the power of a test is small, what has to happen to effect size?

big

What is relability?

is the sample measure an accurate, consistent indicator of the population

What does decreasing variability in the measure do for power?

increase

What is the only real way for researchers to increase power?

sample size

Why is power important when designing a study?

help to decide how many subjects are needed

How do you estimate effect size?

actuarial approach, calculate based on your data, decide what ES is important