1when sufficient cases and a robust design are in place, a _______ analysis is used.
regression
2what type of sample size do you need when you have mutliple predictors
large sample sizes
3what is the advantage of cross-sectional correlational studies
These studies have the advantage of examining the total amount of variance accounted for by multiple predictors and the relative importance of each individual variable as a predictor
4refers to the circumstance in which variables (or, in the case of regression, predictors) are correlated with one another
multicollinearity
5why is it important that multicollinearity be kept to a minimum in regression
that way each predictor can contribute as much unique variance as possible
6how do you determine the unique variance of a predictor
it is entered last into the regression equation
7Due to multicollinearity, the ____ variance will always be less than the variance accounted for in a simple correlation
unique
8two primary types of regression analyses examine multiple predictors:
mutliple linear regression and multiple logistic regression
9includes multiple predictors, and the outcome is a continuous variable (e.g., quality of life), which is analyzed to reveal the total amount of variance accounted for by all of the predictors as well as which predictors are most important
multiple linear regression
10when is multiple logistic regression analysis used
when the outcome is categorical
11how is multiple logistic regression analysis reported
as an odds ratio, risk ratio, or hazards ratio to determine the likelihood that, if one condition occurs, a specific outcome will also occur
12equation for Odds ratio
OR=AD/BC
13in health care research, which types of designs are often used for predictive research
case control designs and cohort studies
14an observational, retrospective, cross sectional study that can answer prognostic research questions concerning which risk factors predict a condition
case control design
15when are case control desgins often used
in epidemiological research conducted after a condition has developed: individuals who already have a condition are matched and compared with individuals without the condition (no random assignment)
16what kind of questions do case control designs answer
prognostic and predictive questions
17what is the landmark example of a case control design
research that compared individuals with lung cancer with individuals without lung cancer and identified smoking as a predictor
18how are cohort studies similar and different from case control designs
similar: both observational
different: participants are followed over time, making this design longitudinal
19the research question is identified before the study begins, and indivuals are followed over time to determine who does and does not develop the condition
prospective cohort study
20Because prospective cohort studies do not use experimental designs, they are weaker in supporting _______________
causative conclusions
21in cohort studies, ___________ are used to express the degree of risk for an outcome over time
risk ratios
relative risk
22existing records or the client's report on past behavior is used to determine if changes occurred over time
retrospective cohort study
23are estimates of risk over time and are typically used when the researcher is interested in reporting time to event analysis
hazards ratio
24in cross sectional research what does the odds ratio describe
the odds of having one condition if the person has another
- does not take time into account
- similar to a correlation coefficient
25in cohort studies that use a longitudinal design, what do risk ratios interpret
the risk of developing one condition if exposed to another, and it takes time into account
26why can't the levels of heirarchy be used for evaluating the quality of descriptive and predictive studies
the lack of random assignment to groups and absence of manipulation of the independent varaible create the opportunity for alternative explanations
27how can you control for alternative explanations
gather data on other predictors and include those alternatives in a regression analysis or odds ratio
28___________ is another strategy to improve the strength of a nonrandomized design
matching
29why is matching a good strategy
mathcing the two groups on important varaibles at the outset will reduce the possibility that differences other than the frouping variable are affecting the outcome
30are prospective or retrospective studies stronger
prospective because individuals can be followed over time
31why are larger samples important
better represent the population and improve the stability of the results
32In a predictive study the relationship between certain potential predictors of a condition can be studied
correlational design
33In a predictive study people with and wihtout a condition can be compared to identify potential predictors
case control design
34In a predictive study individuals can be followed over time to identify differences (predictors) between those who do and don't experience a condition
retrospective and prospective cohort design
35what is the highest level of evidence for a prognostic study
a systematic review that includes two or more prospective cohort studies provides the highest level of evidence for studies that focus on prediction