Quantitative Data Collection & Data Analysis

Question 0

Sampling concepts: population

Answer 0

Well-defined set of people, animals, objects or events

Question 1

Sampling

Answer 1

• involved selecting a group of subjects that are representative of the population under study
• criteria for patient selection is clearly identified before study starts

Question 2

Sampling concepts: target population

Answer 2

Is the entire set of cases the researcher wants to study

Question 3

Sampling concepts: Accessible populations

Answer 3

Represents that part of the target population that could feasibly be included In the study

Question 4

Sampling concepts: legibility criteria

Answer 4

Defined by researcher before study starts

Question 5

Samples and sampling

Answer 5

Sample- is a set of units(elements) that make up the population.
- main thing to look for in a paper “ is the sample representative of the population under study?”

Question 6

Types of sampling

Answer 6

• probability- uses some form of randomization
• non- probability: characteristics chosen by non random means
• generalizability is reduced or limited in non probability sampling compared with probability sampling

Question 7

Non probability sampling: table 12.1 p. 263 (3)

Answer 7

1) convince sample- use of most readily accessible subjects who meet study criteria- has the greatest risk for sample bias
2) quota- knowledge about the population is used to determine the sample
3) purposive- researcher hand picks cases based on his or her knowledge of the population

Question 8

Probability sampling: random selection

Answer 8

Occurs when each subject has equal chance of being assigned to either the control or experimental group
- population has to be identified and then randomization done with random numbers tables, computer program. Etc

Question 9

Stratified random sampling

Answer 9

• population under study divided into subsets that are homogeneous for a particular trait or feature
• a stratified sample would have representation in both the treatment and control group
• a stratified sample is bars on the proportion of subjects in a population

Question 10

Muti- stage sampling or cluster sampling

Answer 10

• successive random sampling of patients that progress from large to small that meet sample eligibility

Question 11

Systematic sampling

Answer 11

• selection of every nth case from a list of subjects or every nth person that walks through a door etc
• issue: bias due to non- randomness can be introduced, not all individuals have equal change of being in the study

Question 12

Matching

Answer 12

• used to construct an equivalent comparison sample group by filling it with subjects who are similar to each subject in another sample group in terms of pre established variables such as age, gender, education, medical diagnosis

Question 13

Sample size

Answer 13

• how do you dried how many patients should be in a study?
• size is determined before study starts
• usually researchers conduct a pilot study to assist in sample size determination
• smaller the effect researcher wants- the larger the sample size needed

Question 14

Sample size

Answer 14

• power analysis- is advanced statistical technique used to identify the number of patients required in a study
• if not done, research studies may be bAsed on samples that are too small
• if the sample Is too small, may lead to a lack of support for the researchers hypothesis or a type 1 error
  Type 1 error- rejecting a null hypothesis when it should hAve been accepted
• therefore, the researcher may say the results are significant when they are not

Question 15

Sample size

Answer 15

• type 2 errors mag also occur- accepting null hypothesis when it should have been rejected if the sample size is too small
• therefore, the sample is too small to detect a difference
• just b/c the sample is large does not ensure no errors will occur if the design of the study is not well thought out

Question 16

Sample size considerations

Answer 16

1. Design used
2. Sample procedures
3. Formula used to estimate ideal sample size
4. Frequency of phenomenon under study
5. Heterogeneity sample
6. Projected cost or money available

Question 17

Critiquing the sampling and sample size

Answer 17

• all research articles will identify the sample selection process in methodology section at beginning
• selection criteria will be stated
• few articles will state the number that was required to identify a small, medium effect. Large effects usually not used

Question 18

Data collection methods

Answer 18

• begins at the same time of the literature review
• researchers look at how other researchers have measured the concept they wish to measure, reliability of the intranets used etc
• must look at nature of research question and the specific population to be studied

Question 19

Operationalization

Answer 19

Variables defined and researcher states how they are going to be observed and measured

Question 20

What to do if not previously measured?

Answer 20

• may need to develop a new tool
• may have to look at current level of evidence and make a choice on the best method to use
• regardless, data collection must be consistent

Question 21

How to keep data collection consistent?

Answer 21

• some method used to administer the tool observation
• research protocol(procedure) very explicit all following the protocol
• often several questions arise and protocol has to be altered to accommodate
• inter-rater reliability- important to end hare consistency between data collectors

Question 22

Physiologic and biologic measurement

Answer 22

• use of specialized equipment that has to be purchased, calibrated prior to data collection, may require more specialized nursing
• includes lab work- some lab, same procedures used?? Same MRI Machine?
• advantage- precise, objective, and sensitive to parameter being measured

Question 23

Observational methods

Answer 23

• are subjects aware of being observed?
• public space verses space where some degree of privacy is expected
• either of being observed and no consent or do not know?
• debriefing subject to let them know have been research subjects and allowing them the opportunity to refuse to have their data included
• observational data collection methods most at risk for research bias

Question 24

Observational methods cont

Answer 24

• field notes- comments made by researchers and may not provide answers for critical areas of focus
• anecdotes: subjects provides a short amount of an event
• standardized tools observers must be trained ( inter rater reliability)
• interviews- structured or unstructured( in person, by telephone) allows subjects in report data directly
• questionaries- open or closed
• closed- usually use a likert scale, defined choices

Question 25

What is best way researchers can determine the stress that new mothers are experiencing in the first month following childbirth?

Answer 25

1) field notes
2) structured interview
3) unstructured interview
4) physiological instrumentation

Question 26

Questionaries

Answer 26

• likert scales very
• problem ceiling effect
• if questionaries too long, patient fatigue, decreased reliability of results, and may not complete if less than 10% of items not answered cannot use

Question 27

Records of available data

Answer 27

• do not have to start off collecting data so saves time and less expensive to do
• names and any identifying info removed
• missing records and missing data may make not representative sample and topic
• authenticity of records
• personal digital assistants collected data ( ideal in clinic)

Question 28

Online computerized methods

Answer 28

• many research now using on-line data collection
• main advantages:
• subjects can do from any location, can access large number of subjects, prevent data entry errors, questionaries complete( no missing data)
• disadvantage: not all subjects have a computer or computer literature