Non-Systems (Research)

Question 1

A physical therapist searches the literature for a reliable and valid questionnaire to measure quality of life in patients with heart disease. The therapist plans to use the questionnaire to assess changes in quality of life before and after participating in a cardiac rehabilitation program. Which questionnaire property is MOST critical to achieve the therapist’s objective?

Answer 1

Responsiveness

Question 2

What is sensitivity?

Answer 2

• Measure of validity of a diagnostic or screening test based on the probability that someone with a disease will test positive on the test.
• Proportion of people with a dz who have a positive test.

Correctly identifies true (+)

SnOut - if you give a highly sensitive test and it comes back negative you can rule out the dz because it is really good at finding (+).

Question 3

What is specificity?

Answer 3

• Measure of validity of a diagnostic or screening test based on the probability that someone who does not have a disease will test negative on the test.
• Proportion of people without a dz who have a negative test.

** Correctly identifies (-) **

SpIn - if highly specific test and get a positive then you rule in the likelihood of dz because this test is really good at finding negatives

Question 4

What is responsiveness?

Answer 4

ability of the instrument to detect change over time. This is an essential attribute when the goal is to assess changes in quality of life before and after participating in cardiac rehabilitation. If the questionnaire is being used to assess the effectiveness of an intervention, the score should change as the patient’s status changes and stay the same if the patient is unchanged.

Question 5

What is linearity of change?

Answer 5

Refers to the magnitude of the change with respect to the starting score. The extent to which a score changes is a function of the starting score. For example, patients may experience substantial changes in joint ROM if their initial ROM is limited because there is more to be gained.

Question 6

Reliability

Answer 6

Reproducibility or repeatability of measurement

Overall consistency of the measurement. This determines the reproducibility of a study.

Question 7

Test-retest reliability

Answer 7

How likely are you going to get the same result on the same test performed again?

Question 8

IntERrater reliability

Answer 8

Consistency or equivalence of measurements made by more than one person.

Indicates the agreement of measurements taken by DIFFERENT examiners

What is the homogeneity of ratings among various researchers?

Question 9

IntRArater reliability

Answer 9

The consistency or equivalence of repeated measurements made by the SAME person over time.

How likely is a single researcher going to get the same result?

Question 10

Validity

Answer 10

Degree to which a useful or meaningful interpretation can be inferred from a measurement.

The ability to come as close the truth as possible

How closely does the result match the known truth or Gold standard test?

Question 11

Internal validity

Answer 11

Focus on cause and effect relationships

Specifically on is there evidence that given a statistical relationship between the independent variable and dependent variable in a experiment, one causes the other.

Question 12

External validity

Answer 12

Extent to which results of a study can be generalized beyond the study sample to persons, settings, and times that are different form those employed in the experimental situation.

Concerned with the usefulness of the information outside the experimental situation

Question 13

Level of evidence

Answer 13

Listed best to least good

1A – Systematic review of RCTs
1B – Individual RCT (narrow confidence interval)

2A – Systematic review of cohort studies
2B – Individual Cohort study
2 C – “Outcomes” research (ecological studies)

3A – Systematic review of case-control studies
3B – Individual case-control study

4 – Case series
5 – Expert opinion

Question 14

Qualitative vs Quantitative

Answer 14

Qualitative - pertaining to a targets behavior or characteristics (case study)….not necessarily testing something more just describing.

Quantitative - numeric analysis of relationships and test results
- statistical analysis
- suggesting correlation

Question 15

Quantitative Studies (Research design)

Answer 15

numeric analysis of relationships and test results
- statistical analysis
- suggesting correlation

• Descriptive design - observation data collection
• Correlation design - observation data collection relates several variables
• Quasi-Experimental Design - Seeks causality (doesn’t manipulate variable)
• Experimental design - classic RCT where the independent variable is manipulated

Question 16

Null Hypothesis

Answer 16

“there is no relationship among the groups” or “any relationship is due to chance”

Most studies try to disprove the null hypothesis

Question 17

Alternate Hypothesis

Answer 17

“there IS a relationship among the groups”

typically what the research is trying to prove

Question 18

Independent vs dependent variable

Answer 18

Independent variable -
Input the characteristic of the person you are looking at

Dependent variable -
Result of the test or measure that is performed

Does does strengthening exercise (independent variable) have any effect on performance on the 6MWT (dependent variable)

Question 19

Level of Measurement scale in research

Answer 19

Nominal - attributes are named
(numbers on soccer jersey, nationality, clinical diagnosis, blood type, occupation)

Ordinal - attributes are ordered
(Borg RPE scale, MMT, pain)

Interval - distance is meaningful. equal distance between numbers with no known zero point
(temperature scale, age)

Ratio - Absolute zero exists. Equal distance with known zero (no negatives)
(weight, number of pts, weight, distance)

Nominal and Ordinal - non parametric
Interval and Ratio - parametric

Question 20

T-test

Answer 20

degree to which data sets are different from each other.
The different between the means/variability of the groups

Question 21

Chi-squared test

Answer 21

assess the goodness of fit between the observed values and expected values.

Question 22

Correlation coefficient

Answer 22

quantifies the strength of relationship between the different variables

(-)1 - (+)1

(-) 1 = negative relationship (walking speed dec. as age increases)
(+) 1 = positive relationship (infant age increases weight increases
0 = no linear relationship

Closer to either end shows the strength of the relationship….closer to 0 is a weaker strength

Question 23

Standard deviation

Answer 23

• represents the amount of variance from the mean

In a normally distributed curve
68% of data points are 1 SD from mean (above or below mean is 34%)
95% of data points are 2 SD from mean
99.7% of data are 3 SD of the mean.

• typically data that is more than 2 SD outside of the norm is considered significant

Question 24

z score

Answer 24

indicates how far an item deviates form the mean in terms of SD

Question 25

Type I Error

Answer 25

Reject null hypothesis when the null hypothesis is true

FALSE POSITIVE

Question 26

Type II Error

Answer 26

Accept the null hypothesis when it is false

FALSE NEGATIVE

Question 27

Minimally Clinically Important Difference (MCID)

Answer 27

smallest change in a test that would be important to a pt.

6MWT MCID is 50 meters

Question 28

Number needed to treat (NNT)

Answer 28

Number of pts who need tx to prevent 1 bad outcome. Ideal number is 1 (so each tx is effective at preventing negative outcomes)

closer to 1 is better (and further away from 1 is worse)

Question 29

What are considered the best/strongest level of evidence?

Answer 29

Systematic reviews!

Systematic review of the randomized control trial is the most.