Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Critical Inquiry > Critical Inquiry Exam 1 > Flashcards

Critical Inquiry Exam 1 Flashcards

1
Q

Evidence-Based Practice (EBP) or Evidence-Based Clinical Decision Making includes what 3 components?

A

Best Evidence
Clinical Experience
Patient Preference

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Nonexperimental Evidence (clinical experience) tends to overestimate or underestimte efficacy?

A

Overestimate

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

For any single study, the most you can say is that the findings for this group, given this procedure, are not likely to have occurred __________.

A

by chance alone

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

The primary aim of a scientific experiment is not to precipitate decisions, but to make an appropriate adjustment in the degree to which one….__________

A

believes the hypothesis being tested.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Background questions refer to _________, while foreground questions relate to_________.

A

general knowledge about a disorder or intervention

specific information that will guide management of the patient

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q 
The clinical question should include the PICO acronym
P:
I:
C:
O:
A

Population: target group, including relevant characteristics
Intervention: What is the treatment?/May also be a prognostic factor or diagnostic test
C: Comparison group (relevant when two or more treatments are being compared)
O: Outcome (what is the outcome of interest - relevant measurements)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Database

A

compilation of topic-related information

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Search engine

A

Retrieval system (typically searches multiple databases simultaneously)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Can you access PubMed at IHP?

A

Yes. At the bottom of IHP webpages, click on library. (library.mghihp.edu)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Searching:

Wild Card

A

helps with different forms of a work (ex: patell*)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Boolean Operators:
AND:
OR:
NOT:

A

AND: both, narrower
OR: either, broader (good for synonyms)
NOT: only one

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Literal Search

A

Conducts search for combination of terms in particular order designated

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Searching:

Quotation Marks

A

Will locate only items contains words “together and in that exact order”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Sparse search results are those with ___ hits and are said to have ___ sensitivity.

A

1-2 hits

low sensitivity

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Excessive search results are those with ___ hits and are said to have ___ sensitivity and ___ specificity.

A

20,000
high sensitivity
low specificity

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Point Estimate

A

The true risk of the outcome in the population is not known and the best we can do is “estimate [of] the true risk based on the sample of patients in the trial.”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

If the confidence interval (CI) is fairly narrow, what does this tell us about the point estimate?

A

If the CI is fairly narrow, then we can be sure that our point estimate is a precise reflection of the population value.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

What must be true for the result to be statistically significant at the 0.05 level?

A

If the result is statistically significant at the 0.05 level, the value corresponding to “no effect” must fall outside the 95% CI.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Questions to ask before you decide to apply the results of the study to your patient:

A

Is my patient so different from those in the study that the results cannot apply?
Is the treatment feasible in my setting?
Will the potential benefits of treatment outweigh the potential harms of treatment for my patient?

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

The relative risk (RR) is defined as:

A

risk of outcome in treatment group/risk of outcome in control group

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

The relative risk tell us _______. RR < 1 tells us the treatment ____ the risk of the outcome. RR > 1 tells us the treatment ___ the risk of the outcome.

A

How many times more likely is in that an event will occur in the treatment group relative to the control group.
RR < 1 tells us the treatment DECREASES the risk of the outcome.
RR > 1 tells us the treatment INCREASES the risk of the outcome.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Absolute Risk Reduction (ARR) or absolute risk difference:

A

risk of outcome in control group - risk of outcome in treatment group

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Relative Risk Reduction (RRR)

A

absolute risk reduction / risk of outcome in control group.

RRR tells us the reduction in the rate of the outcome in the treatment group relative to that in the control group.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Number Needed to Treat (NNT)

A

1 / ARR
Number of patients we need to treat with the experimental therapy in order to prevent 1 bad outcome (incorporates duration of treatment)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

dichotomous outcomes

A

yes or no outcomes that happen or don’t happen

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

“Gold Standard” of experimental trials:

A

Randomized Controlled Trial (RCT)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

Random selection:

Goal?

A

refers to choosing subjects
everyone in the defined population has an equally likely chance of being chosen.
Needn’t be from the entire population but a defined subgroup.
Goal: maximize generalizability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

3 violations of random selection

A

Sample of convenience (all available who meet criteria)
Consecutive/Rolling Recruitment
Purposive Sample (purposefully biased)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

Random Allocation

Goal?

A

Has to do with how subjects get into groups
-should ALWAYS be done except where subjects are grouped by preexisting conditions
Goal: make groups as alike as possible on known and unknown factors

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

2 reasons to have a covariate

A
  1. you think it will affect the relationship between the IV and DV
  2. because it is also of interest but not of prime interest
    Note that adding covariate may create subgroups
    Ex. 1A is female, 1B is male, 2A is female, 2B is male
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

Unocontrolled Variable

A

a variable that you think might affect the relationship between your IV and DV but choose not to assess (too expensive, time consuming, etc.).
You acknowledge the failure to measure it as a limitation to your study.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

Confounder

A

Special case of covariate. The covariate has an independent relationship with both the IV and DV.
May lead you to a false conclusion about the relationship between IV and DV.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
33
Q

Controlled Variable

A

something that you address before the study begins, and control from varying. These variables conceptually vary among people, but will not vary in your study.
Example, choose only females in order to rule out gender as an explanation for the observed difference or association between groups.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
34
Q

Internal Validity

A

Credibility

The degree to which the study minimized alternative explanations (chance, bias, confounding) for the findings

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
35
Q

Compliance Bias

A

People stay in the study but cooperate differently between groups

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
36
Q

Loss to Follow Up (Attrition)

A

People leave the study either during the intervention (attrition) or fail to keep follow-up appointments or submit follow-up questionnaires differently between groups. This could be do to random error, but the concern is that: subjects in one group may be more likely to be compliant or more likely to remain in the study than the other group.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
37
Q

Who experiences the placebo effect?

A

Both groups!

38
Q 
Use RAMMbo to check validity
R:
A:
M:
Mbo:
A

R: Representative (Who did the subjects represent?)
A: Allocation (Was the assignment to treatment randomized? Were the groups similar at the trial’s start?)
M: Maintenance (Were the groups treated equally? Were outcomes ascertained and analyzed for most patients?)
Mbo: Measurements Blinded or Objective (Were patients and clinicians blinded to treatment? OR Were measurements objective and standardized?)

39
Q

4 Criteria to be a Variable:

A
  1. Must potentially vary among subjects
  2. Must be measurable or ascertainable in some way that will yield a value, category, or label
  3. You should be able to assign a value (or level) to each subject in the study
  4. You should be using the value (or level) in some way in an analysis
40
Q

Level/Precision of Variables

rank in order from most to least precise

A
  1. Nominal/categorical
  2. Ordinal
  3. Interval
  4. Ratio
41
Q

Continuous variables can ______.

A

take on any value, including decimals/fractions

42
Q

Discrete Variables are _____.

2 examples:

A

counted as whole unites only

nominal/categorical (eye color, age category, income level, day of week), ordinal

43
Q

Dichotomous Variables are discrete/nominal variables with only ___ levels.

A

2 (male/female, high BP/low BP, acute/chronic)

44
Q

Ordinal Variables (definition and examples)

A

must have order to them, but no conclusions can be drawn about the intervals between numbers and rankings
Ex: rank in class, pain scale (rank 0-5…a pain ranking of 5 is not necessarily 5 times worse than ranking of 1), age category can be used as ordinal data when a number is assigned to each category and an order to the categories is assumed.

45
Q

Interval

A

Interval: Has order and can take on a range of values, but zero point is arbitrary.
Ex: dates and time, temperature, age

Ratio: Has true zero, can take on any value, intervals are equal
Ex: goniometry, isokinetics

generally not considered critical to discriminate between interval and ratio level data

46
Q

As the sample gets more diverse, the ‘error’ of the mean _____ because the mean is less representative of the individuals in the group.

A

increases

47
Q

T/F

Mean Square = Variance = Error = WGV

A

True

48
Q

SD = ?

A

square root of variance

49
Q 
Variance:
The variance (SD^2) gives us a sense of:
A

scatter of scores around the mean

how well the mean value represents the individuals in the group

50
Q

A mean is meaningless without SD because ______.

A

you cannot otherwise know how well the mean represents the group.

51
Q

Normal distribution of the variables (normal or bell-shaped curve) implies that ___% of the individuals fall within 2 SD of the mean value.

A

95%

52
Q

We expect that increasing sample size will _______ the SD because more people make the mean value more representative (less noisy) as long as the new people are drawn form the SAME population.

A

decrease

53
Q

Standard Error of the Mean (SE)

A

an estimate of how much one sample mean may deviate from the actual population mean

54
Q

T/F

Difference between means = effect size = BGV

A

True

55
Q

With less overlap, the difference in two curves are ___ likely to be due to sampling variability (noise).

A

less

56
Q

As BGV increases (presuming fixed WGV), you are ____ sure that the observed differences did not happen by chance alone.

A

more sure

57
Q

How could you make any one sample a better representation of the group (population) from which it is drawn?

A

increase sample size

use more homogeneous groups

58
Q

Width of Curves = WGV
As the curves become narrower (assuming fixed differences between means), there is ___ overlap and we are MORE sure that the two groups are different by more than just chance alone.

A

less

59
Q

Explain the concept of quartiles

A

combines frequency and range; the upper limit of the range of values within which the first, second, third, and fourth 25% of the subjects fall

60
Q

Null Hypothesis

A

You state that there will NOT be an association or difference between the ID and DV

61
Q

Alternate Hypothesis

A

You state that there will be an association or difference between the ID and DV

62
Q

Two-Tailed Hypothesis

Examples:

A

No direction of association or difference is stated

Examples: Null Hypothesis, Alternate Hypothesis

63
Q

One-Tailed Hypothesis

A

The statement of difference (association) or no difference (no association) includes a directional component
Ex: includes phrases such as will not be less than (null), or will be less than (alternate)

64
Q

Does every study state a hypothesis?

A

No, sometimes authors choose to state the study objective or purpose of the study instead. This is fine.

65
Q

How sure you are is given by:

A

1 - pvalue

66
Q

probability of ruling out chance (1-p) and p-value are a function of:

A

Ratio of variability between groups and variability within groups (weighted by sample size and number of groups)

67
Q

T/F

Statistical Significance = Clinical Relevance

A

FALSE

68
Q

What does statistical significance indicate?

A

That random variability is an unlikely explanation for that observed difference or association

69
Q

If the value that reflects the difference between groups is ___ SD beyond zero (no difference or null hypothesis), then chance variation is unlikely to account for observed differences.

A

2

70
Q

By convention, if p>0.05, you will accept the ___, and reject the ____.

A

null hypothesis

alternative hypothesis

71
Q

Alpha Level

A

The level of probability established before the study at which the investigators will reject the null. Alpha determines the level at which you accept or reject the null. Alpha is NOT the p-value actually found in the study.

72
Q

Most common alpha values:

A

5% or 1%

73
Q

Type 1 Error

A

Claim there is a “true” difference when, in fact, there is not (a false positive)

74
Q

You cannot commit a Type 1 error if you do not reject ____.

A

You cannot commit a type 1 error if you do not reject the null hypothesis.

75
Q

A larger alpha increases the likelihood that you will be able to reject the null, and therefore increases the likelihood of committing a type __ error.

A

A larger alpha increases the likelihood that you will be able to reject the null, and therefore increases the likelihood of committing a type 1 error.

76
Q

Lowering the alpha level (increases or decreases) the likelihood of a false positive.

A

Lowering the alpha level (from .05 to .01) decreases the likelihood of a false positive.

Only 1% probability that findings are random variation (not a true difference) instead of a 5% probability.

77
Q

Beta Level

A

The level of risk the investigators are willing to accept in advance that they may have to conclude that observed results are due to chance when in fact, they are not.

78
Q

Type II error

A

probability that a real difference exists, but you couldn’t rule out chance as an explanation. A false negative.

79
Q

You cannot commit a Type II error unless you accept the ___ hypothesis.

A

You cannot commit a Type II error unless you accept the null hypothesis.

80
Q

Lowering the beta level (increases or decreases) the possibility of a false negative.

A

Lowering the beta level decreases the possibility of a false negative.

81
Q

The ____ is the “plumb line” that you use as a reference for your observation.

A

The p-value is the “plumb line” that you use as a reference for your observation.

82
Q

____ is the standard against which you judge the plumb line.

A

Alpha is the standard against which you judge the plumb line.

83
Q

Alpha and beta tend to be ___ proportional.

A

Inversely

84
Q

Consequences of a Type 1 error

A

false positive. If there is a risk to the patient in delivering this treatment, this is an issue!

85
Q

Consequences of a Type II error

A

False negative. You may miss the opportunity to reduce morbidity or mortality.

86
Q

Problem with setting an alpha: It leads to a ___ decision rule.

A

It leads to a dichotomous decision rule – Treatment is effective or it is not.

87
Q

Do authors always specify that they set an alpha level or the value at which they set the p-value?

A

No, but if the authors identify findings as statistically significant, you can use what they report to figure it out.

88
Q

Does an author have to set an alpha level?

A

No!

89
Q

If the authors do not set an alpha level (explicitly or implicitly), then the authors:
Should NOT:
Should NOT:
Should NOT:

A

Should NOT accept or reject the null.
Should NOT talk about Type I or Type II error (based on incorrectly accepting or rejecting the null)
Should NOT discuss Power (based on probability of a Type II error)

90
Q 
Confidence Interval (at 95% probability) indicates:
It can be estimated by:
A

You are 95% sure that the actual value lies somewhere between the upper and lower limits of that range; there is a 5% probability that the actual value lies outside that range.
95% confidence interval can be estimated by calculating to SD around the obtained value.

91
Q

If the null value falls within the 95% CI, the finding (can/can not) be statistically significant at p<0.05.

A

If the null value falls within the 95% CI, the finding CAN NOT be statistically significant at p<0.05.

92
Q

T/F
The mean value (aka point estimate) always falls within the 95% CI, because the CI is constructed AROUND the point estimate.

T/F
Statistical significance is based on the location of the NULL VALUE in the CI, not the location of the point estimate in the CI.

A

True

True

Critical Inquiry (1 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions