Research: lecture 2 Flashcards

(49 cards)

1
Q

cofounding variable

A

another variable that might affect either your IV or DV (unintentional)

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

quantitative

A

testing theories using numbers

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

qualitative

A

testing theories using language

focuses on broad descriptions and understanding complex phenomena without direct manipulation

interview

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

between qualitative and quantitative which one do you control the IV?

A

quantitative

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

single-subject

A

one or few participants are measured many times in order to better understand the process

usually a unique population

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

scientific/alternative hypothesis:

A

statement about the expected outcome or relationship between variables of a study

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

null hypothesis

A

no relationship

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

which hypothesis is usually in the manuscript?

A

scientific hypothesis

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

The ____ of data is one of the key factors affecting the way you analyze the data

A

level

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

what are the different levels of data

A

nominal - naming

ordinal - ordered set with direction

interval/ratio - ordered series of equal sized categories. direction AND magnitude nominal and ordinal

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

nominal and ordinal variables are ____ data

whereas interval/ratio are ___ data

A

qualitative

quantitative

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

“no help, some help, independent” is an example of what kind of data

A

ordinal

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

parameter statistics definition and type of data associated

A

Statistical methods that assume your data follows a specific distribution, usually a normal (bell-curve) distribution.

quantitative data - interval and ratio

need big enough sample size

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

nonparametric statistics definition and type of data

A

Statistical methods that don’t assume a specific distribution — they’re more flexible, or sample size is small

quantitative or qualitative - nominal and ordinal

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

type of data
ROM -
GAIT -
MMT -
Zip codes -
NPRS 0-10 pain scale -

A

ratio
ratio
ordinal
nominal
interval?

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

what are the three types of research studies

A

descriptive
exploratory
experimental

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

descriptive studies

A

describes data, no statistical analyses looking for relationships

retrospective data (previously collected), normative, qualitative

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

exploratory studies

A

looking for relationships between variables

case-control, quasi-experimental, single subject

correlational/predictive/methodological

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

experimental

A

true experimental design with randomization

RCT

also case-control, quasi, single subject

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

T/F: one study alone can prove something

21
Q

goal of a true experiment

A

to demonstrate a cause and effect relationship between two variables (randomization)

22
Q

quasi-experimental

A

do not manipulate IV to differentiate the groups, they use pre-existing participant variables

ex: pre/post op, dx vs dx

23
Q

alpha

A

The cutoff we choose to decide if something is statistically significant

The point which you would consider the result highly unlikely to be by “random” error or coincidence, therefore it must represent meaning or pattern or be significant.

0.05

24
Q

what level is set in advance

25
p value
The actual result from your test — it tells you how likely it is that your results happened by random chance/sampling error if p = 0.0036, then there is a .36% chance that our decision to reject the null hypothesis is wrong (rejected bc p < .05)
26
p< alpha
reject
27
p> alpha
don't reject null
28
what value is determined by the outcome of the statistical analysis?
p-value
29
research validity
the extent to which the conclusions of the research are believable and useful
30
what are the four types of validity
internal validity construct external statistical conclusion
31
internal validity
how confident you can be that the results of a study are due to the intervention or variable being tested - is there evidence that the IV caused a change in dependent | are my methods sound?
32
components of internal validity
-History -Attrition/mortality -Instrumentation -Regression to the mean -Maturation -Repeated testing -Experimenter bias -Selection
33
A ___ is the best design to maximize internal validity
RCT
34
construct validity
are we measuring the construct we think we're measuring? -sometimes it is difficult to define the variables ex: satisfaction and coordination ex: say measuring function and use AROM as a defintion of function, is this demonstrating construct validity
35
external validity
can the results be generalized to my population? -you need to be specific enough to find a difference but not too specific that it is not generalizable ex: if you evaluate a biceps training program in elderly women, you can only generalize to elderly women ex: study include all persons with LBP; gains can be diluted by many others
36
increased _____ can improve generalizability but can result in finding no differences -external validity
diversity
37
statistical conclusion validity
low power -small sample size -excessive variability of DV -violation of statistical assumptions or used wrong test for type of data error rate type 1 and type 2
38
error rate type 1
reject the null when shouldn't -probability (5/100) that you will say no difference when there is
39
error rate type II
fail to reject when should have -usually due to low power/ small sample size -say difference when no difference
40
entire group of individuals of interest
population entire DPT students
41
usually populations are so large that a researcher cant examine the entire group so they select ___ to represent the population in a research study
sample
42
everyone in your population of students in Tx has an equal chance of being selected
simple random sample
43
selects students in a certain order ex/ select every 20th student on list
systemic sampling
44
the sample frame is divided into parts or sections ex/ randomly selects 10 students from every program
stratified sampling
45
the sample frame is divided into parts or section, but only certain parts or sections are used. However, all members of the parts or sections are sampled. Ex/ random students are selected from three randomly selected programs
cluster sampling
46
the members of the sample from volunteer or self select. ex: internet polls, mailed survey that need to be filled out and return.
convenience sampling
47
which is considered the least reliable and therefore the least desirable method of sampling, but easiest!
convenience sampling
48
what is the discrepancy between a sample statistic and its population parameter
sampling error
49
defining and measuring sampling error is a large part of ____
inferential stats