Research Methods Final Exam

Front 1

What term does the state pertain to?

"As randomized, so analyzed"

Back 1

Intent to Treat

Front 2

What is the difference between Relative Risk and Odds Ratio

Back 2

RR: proportion; compares an event in relation to all the events

Odds Ratio: ratio of ratios; compares events with non events

Front 3

What 3 criteria must be present in experimental research?

Back 3

1. cause must precede effect

2. cause and effect must be correlated

3. correlation cant be explained by another variable

Front 4

T/F

Compliant subjects usually have better outcomes than non-compliant subjects regardless of treatment

Back 4

True

Front 5

Describe a Cohort Study

Back 5

Prospective

Front 6

Describe a Case Control Study

Back 6

retrospective

Front 7

Describe Efficacy and Effectiveness in RCTs

Back 7

Efficacy refers to those that received treatment

Effectiveness refers to those who were offered treatment

Efficacy excludes non-compliers

Front 8

Describe internal and external validity

Back 8

Internal: Did it make a difference?

External: Can it be applied

Front 9

Describe the 'last assessment point carried forward' technique

Back 9

Used for non-compliant data, take the last assessment point before they dropped out and carry it forward to each assessment check point

Front 10

What are the two ways of dealing with non-compliers?

Back 10

1. implement strategies to minimize withdrawal

2. report lost data in a flow diagram

Front 11

Describe the psychological theory of the placebo

Back 11

It's not the treatment that works, it's the belief that it will work

Irving Kirsch Anti-depressant study

Front 12

Which study discussed in class emphasizes the subjective nature of pain

Back 12

Talbot, 2000 Wisdom Teeth

Front 13

Describe the process-of-treatment theory

Back 13

showing attention and care for the patient triggers physical reactions that reduce stress and promote healing

Front 14

what is the gold standard study for epidemiology work?

Back 14

Cohort

Front 15

What is the key feature of a cohort study?

Back 15

time

Front 16

What are the 2 types of epidemiological study designs and their subcategories? Define each.

Back 16

Descriptive: Cross Sectional (look at the mean difference)

and Ecological (existing data sets; compare/contrast rate of disease in specific popn)

Analytical: Case-Control

and Cohort (individuals with disease are selected and their exposure is traced back in time)

Front 17

Is a cross sectional design prospective or retrospective?

Back 17

retro.

Front 18

Is a case control design prospective or retrospective?

Back 18

retro.

Front 19

Is a cohort design prospective or retrospective?

Back 19

pro.

Front 20

What is the limitation of the cross sectional design?

Back 20

outcome measured at a single time point

Front 21

What are 3 limitations of an ecological design?

Back 21

Levelof analysis is the population rather than the individual.

Unableto control for the effects of other factors that could obscure the relationship

Only useful at testing initial hypotheses about exposure

Front 22

The Morris et al Occupational Physical Activity Study is an example of __________ study design.

Back 22

Ecological.

Front 23

The Paffenbarger et al. Harvard Alumni study (PA and heart attack levels) is an example of ________ study design.

Back 23

Cohort.

Front 24

The Fish Oil study was a ________ design

Back 24

Cohort.

Front 25

What are the pros and cons of a cohort study?

Back 25

Pros: Temporal sequence, multiple effects

Cons: time, money, loss to follow up, rare disease

Front 26

An odds ratio calculation is used in ______ study

Back 26

case control and cohort

Front 27

incidence rate and relative risk calculations are used in _______ study.

Back 27

cohort

Front 28

What does an odds ratio of 0.30 mean?

Back 28

There is a 70% reduced risk in the disease

Front 29

What are the pros and cons of a case control study?

Back 29

Pros: valid estimate of exposure disease, $

Cons: temporal sequencing not ideal, recall bias, recruit control participants

Front 30

What are the equations for Relative risk and Odds Ratio?

Back 30

RR=[(a/a+b)/(c/c+d)]

Odds Ratio=[(a/c)/(b/d)]

Front 31

What are the 3 criteria for experimental research?

Back 31

1. Cause must precede effect

2. cause and effect must be correlated

3. Correlation cannot be explained by another variable

Front 32

What are the 4 phases of a clinical trial?

Back 32

1. Small number of patients; emphasis on how drug should be given (dose, frequency)

2. Proof of principle: safety and how well it works

3. Large number of patients; compare to the current standard (RCT)

4. Large number of patients; drug already being marketed; Evaluate long term side effects, risks and benefits

Front 33

What is the difference between effectiveness and efficacy?

Back 33

Effective--> on those offered

Efficacy--> on those who receive it

Front 34

What is the difference between internal and external validity?

Back 34

Internal: did the treatment make a difference

External: can it be applied

Front 35

What is the process of treatment theory?

Back 35

Showing attention, and care to a patient triggers a physical reaction in the body that promotes healing

Front 36

What is the Hawthorne effect?

Back 36

Subjects improve their behavior in response to their awareness of being observed

Front 37

Pygmalion effect "self-fulfilling prophecy"

Back 37

the greater the expectation placed upon people, the better they perform

Front 38

What is Relative Risk?

Back 38

the dif btw event rates in relative terms

Front 39

How do we calc RRR

Back 39

ARR/control

Front 40

How do we calc ARR?

Back 40

Treatment-Control

Front 41

What is Absolute Risk?

Back 41

The actual difference in event rates

Front 42

How do we calc NNT?

Back 42

1 or 100/ ARR

Front 43

What is the event rate?

Back 43

the # of people experiencing an event as a proportion of the # of people in the population

Front 44

What are two ways to collect qualitative data?

Back 44

Analytical narrative: Short interpretive description

Narrative vignette: Detailed description that conveys a sense of holisticmeaning

Front 45

What is trustworthiness?

Back 45

quality achieved in a study when the datacollected generally are applicable, consistent, and natural

Is the study competently conducted?

Front 46

What are the 4 techniques for determining trustworthiness?

Back 46

1. Triangulation: need three sources of data

2. Consensual Validation: isthe process of achieving mutual agreement between analysts relative to finalorder of each quote and theme

3. Reliability Check: have someone work backwards—theindependent analyst proceeds sequentially to categorize the quotes into theappropriate lower order levels and themes

4. Data Saturation: occurs when the researcher is nolonger hearing or seeing new information

Front 47

What is the gold standard for measuring PAEE in the free-living environment?

Back 47

Doubly Labelled Water (DLW)

Front 48

What is the most accurate (objective) measurements of PA and Body composition?

Back 48

Actical and iDEXA

Front 49

How do you calc effect size?

Back 49

(Mean of exp-mean of control)/ SD of control

Front 50

Describe the relationship btw effect size, sample size and power

Back 50

as ES increases, N decreases for the same power

Front 51

How does changing the P value influence sample size?

Back 51

Reducing the P value means we need to increase our sample size

Front 52

What does an alpha of .05 mean?

Back 52

there is a 5% chance that the data is due to random variability, and a 95% chance that it is not.

Front 53

What does a power of .8 mean?

Back 53

If the study was conducted again, there is an 80% chance that the same result would be found.

Front 54

What are the equations for eta and omega?

Back 54

Front 55

What is the coefficient of determination?

Back 55

one variable can predict a % of the scores from the other variable

Front 56

Define moderation and mediation

Back 56

Moderation: Underwhat conditions does X relate to Y? (A third variable)

Mediation: Whydoes variable X relate to Y under some conditions but not others?

Front 57

What is a type 1 error

Back 57

reject the null when the hypothesis is True

(There is no difference but we think there is)

Front 58

What is a type 2 error

Back 58

Not reject the null when the hypothesis is False

(There is a difference but we think there isn't)

Front 59

Define effect size.

Back 59

ES is meaningfulness. It deals the with strength and magnitude of the finding

Front 60

Define Power

Back 60

Poweris the probability of rejecting the null hypothesis when the null hypothesis isfalse (e.g., detecting a real difference)

Front 61

Define Coefficient of Determination

Back 61

The coefficient of determinationindicates the portion of total variance in one measure that can be explained,or accounted for, by the variance in the other measure (% of overlap)

Front 62

Advantages (4) and disadvantages (2) of a case-study

Back 62

Advantages

-Emphasis on helping

-In-depth analysis of how the variables affect each other

-More informative

-$

Dis.

-External validity

-Opportunity for bias

Front 63

What does equivocal mean?

Back 63

not clear cut

Front 64

What type of study was used for the Rifle Shooter?

Back 64

Case Study

Front 65

what is the difference btw a case study and a single subject design?

Back 65

Case study uses one person, single subject design uses a small number of people

Front 66

What are the advantages (6) and disadvantages of a single subject design?

Back 66

Advantages

-Focus on measurement of consistency acrosstrials

-Small n

-ID within and btw subject changes

Everyone receives the intervention

-Not hampered by statistical assumptions -Emphasis on socialvalidation

Dis.

Front 67

What is the key feature of a Single Subject Design?

Back 67

ongoing assessment for consistency

Front 68

How do you calculate inter-observer reliability?

Back 68

dividethe smaller total of the dependent variable by one observer by the larger totalrecorded by the second observer and then multiply by 100

Front 69

Describe the 3 types of single subject designs

Back 69

1. Replication Reversal (ABAB): Light switch; turn treatment on and off

2. Multiple Baseline: Baseline gets progressively longer

3. Alternating Treatment: Compares more than one treatment intervention

Front 70

Why is visual analysis of data advantageous?

Back 70

if it’s large enough to see it’s likely to be significant

Front 71

WHat are 3 statistical tests used in Single Subject Designs?

Back 71

1. Split Method: Atrend line is made in the baseline and then extended into the interventionperiod.Comparethe proportion of data points above and below the line across the two phases.

2. Two SD method:calculatevariability around the mean, 2points above the SD in the treatment phase indicates a change in performance.

3. Trend (Time Series Analysis): Compare slopes of line of best fits

Front 72

Epidemiological research is _________ not _______

Back 72

Observational not experimental

Front 73

Epidemiology research is used to provide the scientificbackbone for public health endeavors, including:

Back 73

Magnitude of the problem

risk factors

allocate resources

monitor prevention strategies

Front 74

Descriptive vs Analytical

Back 74

D: general observations of disease relationship to demographics

A: test specific hypotheses using purely observational methods

Front 75

What study was the first time epidemiology was used to link levels of PA to disease?

Back 75

Harvard Alumni, Paffenbarger et al

Examining activity levels and rates of heart attacks

Front 76

Clinical Trial

Back 76

test whether one health care interventionis superior to another

Front 77

Selection Bias

Back 77

Occurswhen groups are formed on some basis other than random assignment. Thus, whentreatments are administered, because the groups were different to begin with-->rivalhypothesis that differences are due to the initial selection biasesrather than treatment

Front 78

How do we deal with withdrawl in an ITT analysis

Back 78

- Designtrial to minimize withdrawal

-Accuratelyreport withdrawal

-Analyticapproaches (lastobservation carried forward, other imputation techniques) canbe used to reduce, but not remove the effect of withdrawal

Front 79

Hawthorne effect

Back 79

Feeling of being studied=perform better

Front 80

Pygmalion Effect

Back 80

greater the expectation placed upon people, the better they perform

Front 81

Expectancy Bias

Back 81

refersto the experimenters’ or researchers’ anticipating that certain participantswill perform better

Front 82

How do you calc RRR

Back 82

1-(treated/untreated)

Or ARR/control

Front 83

How do you calc ARR

Back 83

treated-untreated

Front 84

How do you calc NNT?

Back 84

1 or 100/ ARR

Front 85

Statistical Sign.

Clinical Sign.

Back 85

Stat: is this difference likely to be real (not due to chance)

Clinic: is this difference likely to meet the measurement criteria