Brainscape's Knowledge GenomeTM

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


See full index

Research Methods in Health > Week 05 > Flashcards

Week 05 Flashcards

1
Q

Meta means…

A

beyond

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

What does a meta-analysis do….?

A

summarises results, improves precision of estimates of effects in population

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

Do all systematic reviews have meta-analyses?

A

No

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

______ review is the traditional literature review where author selects, reads and summarises without systematic process

A

narrative

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

A narrative review has risk of ___

A

bias (unreliability)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Which is more reliable, narrative or systematic review?

A

systematic

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

What are some disadvantages of systematic reviews/meta analysis ?

A
  • Cant review information that isn’t there
  • Can’t improve on poor quality information
  • May be difficult to validly combine statistical results from studies with different methods
  • Many systematic reviews give open finding (inconclusive) because sufficient high quality studies are lacking
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

A ____________ is a statistical method of combining results from multiple studies

A

meta-analysis

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

“pooled effect size” is a _____ result combining samples from multiple studies

A

single

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

Do larger or smaller samples provide more precise estimates?

A

Larger (more information)

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

List reporting biases affecting whether research findings are published at all (4)?

A
  • Publication bias: unfashionable research topics may miss out on publication
  • Time lag bias: rapid or delayed publication of results
  • Duplicate publication bias: the same research findings published in more than one article; also known as “salami slicing”
  • Outcome reporting bias: marked preference for studies with “significant” results; non-significant may be just as important
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

List biases affecting whether report included in review (4)?

A
  • Citation bias: whether the report appears in a reference list affects its chance of being spotted
  • Database inclusion bias: article not in database is harder to find
  • Language bias:
    preferences for studies in english
  • Personal biases of researchers doing the review
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Funnel plot shows effect size by ____ size

A

sample

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

Bias against trials with small samples and negative/positive effects?

A

negative

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

If no publication bias, funnel plot will look symmetrical or asymmetrical?

A

symmetrical

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

Heterogeneity = ??

A

difference or diversity

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

Can clinical heterogeneity can cause statistical heterogeneity ?

A

Yes

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

Appraising systematic reviews: Validity of the methods used in the review

A
  • Clearly focussed question with clear eligibility criteria
  • High quality, relevant studies
  • Assessment of bias in the studies
  • Results combined across studies, but only if reasonable to do so
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Appraising systematic reviews: Results of the review

A
  • Overall results of the review are presented

- Results are sufficiently precise

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

Appraising systematic reviews: Relevance of the results to your clinical practice

A
  • Outcome considered from the practitioners perspective

- Applicability of results to your own target population

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

What is ecological fallacy?

A

“what applies to the group, will also apply to the individual”

  • not sound concept
  • Internal validity threat
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Value of systematic reviews to practitioners depend on (2)?

A
  • Availability and quality of primary studies for review

- Methods used to conduct the review

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

Meta-analysis estimates overall statistical effect size

A
  • Effectively increases the sample size, leading to higher precision
  • Feasibility affected by heterogeneity among studies
  • Can identify bias in individual studies
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Standard review process applies to quantitative research

A
  • Strong emphasis on clinical trials but observation studies eligible
  • Qualitative research developing alternatives - meta synthesis
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Systematic reviews can/can’t be affected by same biases as primary studies and biases specific to systematic reviews (cannot improve on original evidence quality –> heterogeneity is a risk)

A

can

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

Accidental, chance differences between same means and population average can lead to the risk of the researchers wrongly believing that sample accurate for whole population —> this is where ___ ____ comes in

A

inferential statistics

27
Q

Rarely possible to get complete data from population (impractical, extensive, instead researchers rely on _____ drawn from the population—> therefore, most quantitative research is at risk of sampling error)

A

samples

28
Q

Descriptive statistics describe the sample….?

A

mean, mode, median, SD, min, max, range amount etc

29
Q

How to sample results generalise beyond the sample affects the ___ validity

A

external

30
Q

Population parameters (statistics) =

A

measurement or results for populations (means, SD, median numbers ratio etc)

31
Q

Inference =

A

conclusion based on information (see smoke –> information, conclude there is a fire –> inference)

32
Q

Inferential statistics enable conclusions (inferences) about _____ when only _____ data is available

A

populations, sample

33
Q

Inferential statistics deal with ______ information

A

uncertain

34
Q

______ is a prediction about what a result will be before the data is analysed or even collected

A

hypothesis

35
Q

Hypotheses are exploratory or descriptive?

A

descriptive

36
Q

Choosing hypotheses after inspecting the data is bad science… is like

A

asking a question after finding the answer

37
Q

Are hypotheses proved and disproved?

A

No, they are either consistent or inconsistent

38
Q

Null hypothesis is _____ the hypothesis

A

against

39
Q

Null hypothesis is called H___

A

H0 (against hypothesis)

40
Q

Alternative hypothesis is ___ the hypothesis

A

for

41
Q

Alternative hypothesis is H___

A

H1 (For hypothesis)

42
Q

Hypotheses are usually state or implied?

A

Implied

43
Q

A _____ hypothesis is something that does not happen, no effect, treatment does not work, effect size = 0 (no difference between groups)

A

null (H0)

44
Q

A _____ hypothesis is something that does happen, is effect, treatment works, effect size is NOT 0 (there IS a difference between groups)

A

alternative

45
Q

Is a non-zero effect the same as “treatment works”… explain why

A

No, because it does not state the direction (could harm or benefit pt, AND doesn’t account for minimal significance)

46
Q

Equivalence is ________ hypothesis

A

nondirectional

47
Q

Equivalence/nondirectional hypothesis are ____ interesting because we want to find benefit but the default statistical significance test

A

rarely

48
Q

Superiority - directional hypothesis, H0 & H1:

A
  • H0: treatment is not superior = does not have a larger beneficial effect than control
  • H1: Treatment is superior = does have a larger beneficial effect than control
49
Q

Equivalence - nondirectional hypothesis, H0 & H1:

A
  • H0: treatment and control have equal effects, no difference
  • H1: Treatment and control have non-equal effects, a difference in either direction
  • H1: goes either way, treatment better OR worse than control
50
Q

Non-inferiority - directional hypothesis, H0 and H1:

A
  • H0: treatment is “not unacceptably worse” than control
  • H1: treatment is unacceptably worse than control
    (common but questionable for EBP)
51
Q

Hypotheses are about ______, not samples

A

populations

52
Q

Hypotheses are for making ______ about populations based on sample data

A

inferences

53
Q

Inference is a conclusion (about the ______) based on information (from the ____)

A

population, sample

54
Q

Method to test inferences is ___ _____ that use sample data to infer beyond sample

A

inferential statistics

55
Q

Inferences from sample to population are _____ –> thats why inferential statistics deals with it

A

uncertain

56
Q

Qualitative research typically ____ hypotheses

A

avoids

57
Q

H0 null hypothesis goes with statistical non-significance p ___ .05

A

> (greater than)

- zero effect

58
Q

H1 alternate hypothesis goes with statistical significance p ___ .05

A

< (less than)

- non zero

59
Q

If p < .05, ____ than 5% chance of sample effect if H0 is true

A

less (unlikely data)

60
Q

If p < .05, ____ than 5% chance of sample effect if H0 is true

A

more (likely data)

61
Q

Type 1 error…?

A

H0 is falsely rejected when p < .05

- accidental error not calculation error

62
Q

Type 2 error…?

A

H0 is falsely retained when p > .05

63
Q

Limitations to hypothesis testing

A
  • Clinicians want to know HOW WELL intervention works, not whether effect is zero
  • H1 “non zero” not informative for EBP (if rejected, all we know is data is unlikely true, what use is that?)

Research Methods in Health (12 decks)

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