Module 2

Question 1

What are confounding variables

Answer 1

things that confound information between variables (fiber study = vegetarians could have high fiber diets compared to normal diets = confounding variable)

Question 2

How do you control for confounding variables

Answer 2

research design

Randomization - theoretically, will solve
Homogeneity = keep out vegetarians
matching = for each man in one group have one in 2nd group etc.

Statistically

Question 3

Blinding

Answer 3

researches/participants do not know who got intervention and who did not

Question 4

crosssectional data collection

Answer 4

collecting data at one point in time

Question 5

longitudinal data collection

Answer 5

collect data over period of time

Question 6

prospective

Answer 6

looking forward

Question 7

retrospective

Answer 7

looking backward

Question 8

3 classifcations of quantitative research

Answer 8

experimental
quasi-experimental
nonexperimental

Question 9

Experimental design arrangement

Answer 9

Intervention
control
randomization
experimental group and control group

Question 10

Quasi-experimental design arrangement

Answer 10

Intervention
Control

NO RANDOMIZATION

Question 11

Types of experimental designs =

Answer 11

post-test only R X O
R O

Pretest - post test design R O X O
R O O

Crossover design R O O O X O
R O X O O O

Question 12

Advantages and disadvantages of experimental designs =

Answer 12

Advantage = testing cause and effect relationships

Disadvantages = control vs practical significance (doesn’t reflect real life or practice)
generalizability
need to randomize participants

Question 13

Quasi-experimental design examples =

Answer 13

nonequivalent control group post-test only design
X O
O

Nonequivalent control group pretest posts test
O X O
O O

One group pretest post test design
O X O

Time series design OOOXOOO

Question 14

Quasi-experimental design advantages and disadvantages

Answer 14

advantages = practical, more acceptable to participants (they can select which group they want to be in), Can use when it’s unethical to randomize participants

Disadvantages = more difficult to make causal inferences

Question 15

Nonexperimental designs

Answer 15

correlational and descriptive studies

Question 16

Correlational studies =

Answer 16

study relationship between variables that are NOT MANIPULATED; correlation does not prove causation

Question 17

Correlational study types

Answer 17

cohort = prospective or retrospective longitudinal study look at exposure of two groups and follow over time
case-control = look back at what exposures they had

Question 18

Nonexperimental design advantage/disadvantage

Answer 18

advantage = very practical, efficient way to collect large amounts of data

disadvantages = cannot make causal inferences, self-selection

Question 19

Critiquing a quantitative study things to think about

Answer 19

-statistical conclusion validity = power/sample size, attrition (enrolled 100 but only 50 completed)
- external validity = generalizable to the outside population? (sample plan & scope)
- construct validity = looks at interventions and conditions of study and if they are real representations of the constructs they are trying to study

Question 20

What is internal validity =

Answer 20

Look at whether there is a relationship between the independent and dependent variables. Did the independent variable manipulation really cause the change in the dependent variable?

Question 21

What is a way to increase internal validity

Answer 21

randomization

Question 22

Threats to internal validity

Answer 22

temporal ambiguity
selection
history
maturation
mortality/attrition

Question 23

Temporal Ambiguity =

Answer 23

Does the cause really precede the effect?

This is not a problem in randomized control trials because the researchers create or introduce the independent variable and then observe what happens.

Question 24

History Threat

Answer 24

occurrence of events concurrent with the independent variable that can affect the outcome.

Popstar condom campaign during sex education intervention

Question 25

Maturation Threat

Answer 25

arises from processes occurring as a result of time (growth, fatigue, Etc.)

Question 26

Mortality/Attrition Threat =

Answer 26

people dropping out of study being lopsided

Question 27

Population vs accessible population

Answer 27

population = all female soccer players accessible = who we can actually study (female soccer players in Cincinnati)

Question 28

nonprobability sampling

Answer 28

nonprobability = convenience sampling (flyers in a hospital), quota sampling (50 intramural soccer players and 50 professionals), consecutive sampling (first come, first serve - enrolling first 50 people), purposive sampling (enrolling who you want by selection)

Question 29

Probability sampling =

Answer 29

subjects are randomly enrolled from the total group Simple = list of every soccer player in the world and randomly choose who is enrolled. Stratified probability = random samples from each group Systematic sampling = total group list and take every 10th person

Question 30

Power analysis =

Answer 30

tells you how many people you need to meet in study

Question 31

Things to think about when critiquing sample and setting

Answer 31

Is there a clear description of the sample strategy provided? Is this a representative sample? Is there an appropriate number of subjects? Is there a clear description of the setting? Is the setting representative of where care would be provided?

Question 32

What are levels of measurement

Answer 32

nominal - categorical with no order (race, gender, ordinal - categorical WITH ORDER (education level, Likert scale interval - continuous data (height, weight, etc.)

Question 33

Reliability =

Answer 33

the consistency that an instrument measures something Stability = measures stay the same internal consistency = whether all items on equivalence = interrelation reliability

Question 34

Validity =

Answer 34

are results valid face validity = content validity = appropriate sample for content criterion-related validity = does score relate to some external criteria Construct validity = testing in known groups (give tool to really intelligent and nonintelligent to see if tool

Question 35

Relation between reliability and validity

Answer 35

Can have a tool that is reliable but NOT valid (scale giving the wrong measurement consistently) If tool is VALID, it will always be reliable

Question 36

Different ways to collect data

Answer 36

existing records = healthcare retrospective data original data = self-report, observation (categories, rating scales, observational sampling), biophysiological measures (in vivo - BP, HR, in vitro - lab samples)

Question 37

Things to think about with self-reported data

Answer 37

was it an interview or questionnaire (people can skip question in quesstionnaire but may be more honest) open-ended vs close-ended questions investigator developed vs validated tool

Question 38

Things to think about when critiquing measurement and data collection plan

Answer 38

can I trust the data? Do the data accurately and validly reflect variables under study? Is there reliability and validity of measures reported? Is there a clear description of the data collection procedures provided? Was the best method of data collection used? Were efforts made to enhance the data quality?

Question 39

What are descriptive statistics

Answer 39

Frequency distribution = charts, look at skew central tendency = mode(most frequently occurring), median (exact middle of data), mean (average) measures of variability = range (highest score subtracting lowest score) and standard deviation (what is the variation?)

Question 40

What is inferential statistics =

Answer 40

hypothesis testing (null, research hypothesis, level of significance, statistical test used)

Question 41

inferential statistic tests

Answer 41

t - test independent or paired ANOVA = three or more groups correlation regression = prediction