Brainscape's Knowledge GenomeTM

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


See full index

CU: Research Methods > Final Exam Material > Flashcards

Final Exam Material Flashcards

1
Q

What is the definition of statistics? What are the two classifications?

A

Procedures for collecting, analyzing, interpreting, and presenting data
1) Descriptive Statistics
2) Inferential Statistics

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

What is the definition of descriptive statistics? An example?

A

Used when measuring characteristics of a group without intending to generalize beyond the group
Example: Mean +_ SD: age, BMI, height

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

What is the definition of inferential statistics? An example?

A

Used when making generalization s or inferences from a smaller group (sample) to a larger group (population)
Examples: t-test or ANOVA, correlations: Person’s Correlation Coefficient, Simple Linear Regression, & Multiple Regression

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

What is the definition of a population? What is the value representing a characteristic of the population?

A

A large group to which the results of a study conducted on a sample from the group may be generalized
Value: parameter

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

What is the definition of a sample? What is the value representing a characteristic of the sample?

A

Individuals from the population who actually participate in the research –> a representative subset of the population
Value: statistic

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

What is a sampling frame?

A

A list of all those within a population who can be sampled (sample is taken from this)

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

What are the steps in sampling?

A
  1. Define the population by specifying the criteria (inclusion and exclusion) for selecting participants
  2. Develop the plan for sample selection
  3. Determine the sample size (power analysis?)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What are the options for sample selection?

A

Random Sampling
- Simple Random Sampling
- Systematic Random Sampling
- Cluster Random Sampling
- Stratified Random Sampling
Non-Random Sampling
- Convenience (Volunteer) Sampling

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

What is the definition, advantage, and disadvantage of simple random sampling?

A

Every person in the population has an equal chance of being selected for the sample
- Advantage: unbiased, representative sample
- Disadvantage: Need a good sampling frame

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

What is the definition, advantage, and disadvantage of systematic random sampling?

A

Every nth person from sampling frame is selected to participate
- Advantage: easier to administer than simple random sampling
- Disadvantage: may be biased if pattern in population

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

What is the definition, advantage, and disadvantage of cluster random sampling?

A

Specific clusters (groups) are randomly selected out of all possible clusters
- Advantages: time- and cost-effective (convenient and practical), large samples
- Disadvantages: May be biased/non-representative if clusters are different from each other

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

What is the definition, advantage, and disadvantage of stratified random sampling?

A

Members of the sampling frame are divided into subgroups (strata)
- Random sample of participants then selected from each strata
- Advantages: Good representative sample: captures key characteristics of the population
- Disadvantages: Difficult to administer, need detailed information about the population

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

What is the definition, advantage, and disadvantage of convenience sampling?

A

Investigators recruit easily available individuals who meet criteria until meet desired sample size
- Advantage: very convenient
- Disadvantage: sample may be biased, self-selection bias

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

What type of sample selection is a survey of admissions of individuals with Type II diabetes to hospitals in greater Dayton area?

A

Cluster

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

What type of sample selection is a study of risk factors for progression of osteoarthritis (assuming more women affected than men)?

A

Stratified

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

What type of sample selection is a study of rural access to health care in the different regions of Ohio?

A

cluster

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

What type of sample selection is a survey of of satisfaction and retention among PA in Ohio?

A

Simple or systematic

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

What type of sample selection is a study of the effectiveness of resistance training on functional outcomes in patients with chronic obstructive pulmonary disease?

A

Convenience

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

What is the difference between random selection vs random assignment?

A

RS: about how elected: more representative more generalizable
RA: making groups: more equal = good internal validity, better for causality

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

What are the types of frequency distributions?

A
  • Normal distribution: symmetrical bell-shaped curve
  • Non-Normal distribution: highest frequencies of scores do not fall centrally but are shifted towards positive or negative extremes
    (positively, negatively skewed: where the tail is)
  • Kurtosis: a vertical shift in the normal curve; the middle of the curve is elevated or flattened
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What are the two aspects of descriptive statistics?

A

Measures of central tendency: mean, median, mode
Measures of variability: range, standard deviation, variance, standard error of the mean (SEM)

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

What are the measures of central tendency?

A

Extent to which values cluster in a data distribution plot
- Mean: the average of all the number
- Median: middle number in the list
- Mode: most frequently occurring number
Normal distributions: use mean
Skewed distributions: use median

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

What are the arrangement of skewed distributions?

A

Positively: mode < median < mean
Negatively: mean < median < mode

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

What are the measures of variability? (Elaborate on the first two)

A

How the scores vary; how they are dispersed around the measures of central tendency
- Range: the difference between the highest and lowest scores (ie age range)
- Standard deviation: a numerical indicator of the spread of values within a data set
- Variance
- Standard Error of the Mean

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

What could a large standard deviation around the mean indicate?

A
  • wide spread of data
  • indicate outliers
  • sample size too small
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

What is variance?

A

The square of the standard deviation; used to calculate many other statistics

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

What is the standard error of the mean? What makes it smaller and what does this mean?

A

an estimate of the expected difference between the sample mean and the population mean. (SEM = SD/ square root of n)
Smaller standard deviation, larger sample size the smaller the SEM
The smaller the SEM the greater the confidence that the sample mean accurately represents the population mean.

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

What is the sampling error?

A

The difference between a calculated sample mean and the (unknown) population mean

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

Which hypothesis is typically tested in inferential statistics?

A

The null hypothesis

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

If the level of significance is 0.05 what does this mean for the chance that the study results are erroneous? What does it mean for the researcher’s confidence?

A

5% chance of the study results being in error
95% confident they will detect a true difference when there is one

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

When is a alpha level of 0.01 or 0.001 selected?

A

Medical interventions, pharmacutical companies

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

What is the p value?

A

The probability value–calculated by the computer. The probability of the results being due to chance–measure of the strength of evidence against the null hypothesis.
Low: strong evidence against null
High: weak evidence against null

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

In clinical studies what are the concerns of type I and type II errors?

A

Type I: pts treatment does not actually work
Type II: treatment was effective but was rejected and is no longer available

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

What is the B level typically?

A

0.2 (20%)

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

What is the most common reason for a type II error?

A

small sample size causing low power

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

What is the trade off of level of significance and type I and II errors?

A

Lower alpha decreases risk of Type I error and increases risk of Type II.

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

What is the power of a study? What does it mean if it is high? What is the equation? What is the typical level in health research? What kind of sample size is important for adequate power?

A

The probability of making a correct decision regarding the null hypothesis.
- High: will correctly detect statistically significant differences when they do occur
- Power = 1 - B
- Typically 0.8 (80%)
- Large enough sample size

38
Q

What are the assumption s of parametric statistics?

A
  • normal distribution
  • homogeneity of variance
  • interval or ratio (continuous) data
39
Q

When are non-parametric statistics used?

A
  • not normally distributed
  • small sample size
  • nominal or ordinal level data
40
Q

What is Levene’s Test for Equality of Variances?

A

if the p value is > 0.05 the variances are not significantly different and equal variances are assumed

41
Q

What is a t Test used for?

A

To compare two groups on one dependent variable

42
Q

What is ANOVA and what is it used for?

A

Parametric statistic used to compare the means of
- more than 2 groups
- 2 groups at multiple points in time
- 1 group with 3 or more time points

(see if they are different or not)

43
Q

What are the ANOVA assumptions?

A
  • normal distribution
  • homogeneity of variance
  • interval or ratio (continuous) scales
44
Q

What is the One-way ANOVA?

A
  • One independent variable/factor
  • Three or more levels (groups)
  • Testing the null hypothesis (no difference)
45
Q

What is the statistic calculated from an ANOVA? What does it represent?

What about the p value?

A

Statistic: F ratio
Between group variance (b/c treatment variable) / Within group variance (b/c error)
p value: chance probability of F ratio

46
Q

What is the Post Hoc test? Why is it used?

A

Determines where differences exist.
Used if the one-way ANOVA produces a significant F-ratio (at least one group is different)

47
Q

What is the Two-Way (factorial) ANOVA?

A

Study effect of more than one independent variable on a dependent variable
(2x2) determines:
1) Main effect of Variable A alone
2) Main effect of Variable B alone
3) Interaction (joint) effect of Variable A and B

48
Q

What are the plots of a Two-Way (factorial) ANOVA?

A

No interaction effect: parallel (near parallel) lines
Interaction effect: Nonparallel or intersecting lines

49
Q

What are repeated measures ANOVAs used for? Note?

A

1) changes in mean scores over: three or more time points
2) differences in mean scores under: three or more different conditions

*Present conditions in random order to prevent order effects (practice/fatigue)

50
Q

What is correlational research? (category, investigation of what & how shown, questions)

A
  • type of descriptive research
  • systematic investigation of relationships (correlation/association) among two or more variables
  • “Do the variables change together in a predictable way?”
  • scatter plots show the variable relationship

Person Correlation coefficient, simple linear regression, multiple regression

51
Q

What is a positive correlation?

A

One variable increases as the other variable increases

52
Q

What is a negative correlation?

A

One variable increases as the other variable decreases

53
Q

How is the strength of a correlation determined?

A

Line of best fit: degree to which the values in a scatter plot cluster around a line running through them
from -1.0 to +1.0

54
Q

What does finding a correlation not mean?

A

causation
only that they are related

55
Q

What is the most common parametric correlation statistic?

A

Person product-moment correlation coefficient (Person’s r)
for the relationship between two variables

56
Q

What are the assumptions for a Pearson correlation?

A
  • normal distribution
  • linear relationship
  • interval or ratio (continuous) level
  • no outliers
57
Q

What are the guidelines for interpretating Person’s r values?

A

0.8 and > Very strong
0.6-0.8 Strong
0.4-0.6 Moderate
0.2-0.4 Weak
0.0-0.2 Very Weak

58
Q

How is the significance of a correlation coefficient determined?

A

By a test of significance (random error or good estimate)
null: there is no correlation, correlation coefficient = 0

*Significant correlation does NOT necessarily mean there is a strong correlation just that the results are not likely due to chance

59
Q

When is analysis of covariance (ANCOVA) used as a statistical test in research studies? (skip)

A

When it is known that the group means are different at the beginning of an experiment.

60
Q

What is the nominal level of measurement?

A

Lowest level: Categories by name or label

Example: Gender, Handedness

61
Q

How is the ordinal level of measurement different from the nominal level? Give two examples of the ordinal level of measurement.

A

It is also non-numerical data but it can be put into an order (spacing may not be equal) Ordered Ranking

Examples: Likert Scales (attitudes, preferences), Manual muscle testing (Graded from 0-5)

62
Q

How is the interval level of measurement different from the ordinal level? Give one example of the interval level of measurement.

A

Numeric measures with ordered rankings of equal distances/intervals between units of measurement

Example: Temperature scale

63
Q

How is the ratio (continuous) level of measurement different from the interval level? Give two examples of the ratio (continuous) level of measurement.

A

Numeric measures with ordered rankings, equal spacing between adjacent numbers and a true zero

Examples: walking speed, height

64
Q

Which level of measurement tends to provide more information and allow for more sophisticated statistical analyses than the other levels of measurement?

A

Ratio (continuous) measurement

65
Q

Independent Variable:

A

Cause - variable that the researcher manipulates; the experimental treatment

66
Q

Dependent Variable:

A

Effect - variable being measured to determine the effects of the independent variable

67
Q

What is experimental research designed to do?

A

establish a cause-effect relationship between an independent variable and a dependent variable

68
Q

What are the three classifications of experimental research?

A

1) True Experimental
2) Quasi Experimental
3) Pre-experimental

69
Q

What are the three essential characteristics of true experimental research?

A

1) Manipulation of an independent variable
2) Control (alternative) groups
3) Random Assignments to Groups

70
Q

What is the purpose of random assignments to groups?

A
  • Each participant has an equal chance of being assigned to any group
  • Increases the likelihood that the groups will be equivalent on key characteristics at baseline
  • Controls for most threats to internal validity.
  • Researchers must still do best to minimize the threats of history, instrumentation, and mortality
71
Q

What are the types of true experiment designs and their symbols?

A

Control group: G1: R D1 T D2 G2: R D1 D2
Two Group: G1: R D1 T1 D2 G2: R D1 T2 D2
Multi-Group: G1: R D1 T1 D2 G2: R D1 T2 D2 G3: R D1 D2

72
Q

Why do some True Experiments have only a post-test design? What’s the concern?

A
  • Want to know response to something can only test for it after have intervention
  • Concern: Cannot do a statistical test to affirm group equivalency
73
Q

What are the characteristics of Quasi-Experimental Research Designs?

A
  1. No random assignment (already in groups)
  2. Control Group OR multiple measures
74
Q

What are the types of Quasi-Experimental Research Designs?

A
  1. Nonequivalent Control Group Design
  2. Repeated Measures Design
  3. Time Series Design
75
Q

What is the symbol pattern, use, and major threat/minimization strategy to internal validity for nonequivalent control group design?

A

G1: D1 T1 D2 G2: D1 T2 D2
- Use: When person-factors (IQ, personality, habits) are the independent variable: not ethical or feasible to manipulate
- Threat: Selection Bias
- Minimize: match groups by relevant variables and statistically compare

76
Q

What is the symbol pattern, use, and major threat/minimization strategy to internal validity for repeated measures design?

A

T1–D1–T2–D2–T3–D3 (each person becomes own control)
- Use: small sample size
- Testing
- Minimized: allow for adequate rest between exercises, randomized order of tests

77
Q

What is the symbol pattern, use, and 2 major threats/minimization strategy to internal validity for Time Series Design?

A

D1 D2 D3 {Pre-test} T D4 D5 D6 {Post-test}
- Use: See change overtime, effect of policies or rules
- Threat: History & instrumentation
- Hard to control for

78
Q

How do quasi experiments relate to external/internal validity? Why?

A

Tend to have stronger external validity (occurs in real-world situations)
Tends to have weaker internal validity (due to lack of randomization)

79
Q

What are strategies to reduce internal validity threats in Quazi-Experimental tests.

A

1) Matched Groups
2) Test (statistically) for group equivalence
3) Larger groups

80
Q

What does simple linear regression do?

A

Estimates how variable X (predictor variable) predicts variable Y (criterion variable)

81
Q

What is r2 (squared)?

A

Gives indication of accuracy of prediction of simple linear regression
- How much of the variance in Y is accounted for by knowing X

82
Q

What is multiple regression used for?

A

To predict a criterion (dependent variable) from two or more predictor variables (independent variables)

83
Q

What is multiple correlation?

A

R: calculated as correlation of two or more predictor variables with criterion variable.

84
Q

What is R2 (squared)?

A

the effect of each variable of a multiple regression in explaining variance in the criterion

85
Q

What questions must be answered to select an appropriate statistical test?

A
  1. Is the research question focused on a relationship between variables or a difference between groups (cause-effect)?
  2. How many groups of participants are involved?
  3. If there are two groups only, are they independent of each other or related?
  4. How many independent variables make up the study?
  5. Are there multiple measures of the dependent variable?
86
Q

What is a common principle for dealing with drop outs?

A

Intention to treat. Evaluate all data even if drop outs happen

87
Q

What does clinical meaningfulness mean? How is it assessed?

A

The practical importance of a treatment effect; whether it has a noticeable effect on daily life
Assessed by:
- effect size
- minimal detectable change (MDC)
- minimal clinically important difference (MCID)

88
Q

What is effect size? (calculated from, guidelines)

A
  • strength of a treatment effect
  • indication of the meaningfulness of results

Calculated from sample means and standard deviations
> 0.8 Large
0.5-0.8 Moderate
0.2-0.5 Small
< 0.2 weak

89
Q

What is MDC?

A

Minimal detectable change
smallest change in a score on an outcome measure that likely reflects true change

Calculated using statistical data, including SEM

90
Q

What is MCID?

A

Minimal Clinically Important Difference
smallest change in score on an outcome measure that is perceived to be beneficial by the patient

Calculated by different methods: (SD, SEM, ES)

CU: Research Methods (8 decks)

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