Module 5: Quantitative research design

What are the 2 types of quantitative research study designs?

• experimental

• non-experimental

What are the 2 types of experimental designs?

1. randomized

2. non-randomized (quasi-experimental)

Does RCT have a control group?

yes

Does quasi-experimental have a control group?

it could depending on experiment

What is an intervention?

when researcher does something to subject

• introducing intervention/tx

What is a control?

a group that does “nothing”

What is randomization?

researcher assigns participants to a control or experimental condition on a random basis

What type of quantitative design is the gold standard?

RCT

Describe what a control group may look like

• placebo

• alternative tx

• standard of care

What is allocation concealment?

process of making randomization sequence unknown

Why is allocation concealment important?

ensures randomization reminds valid until randomization

What type of experiment is blinding or masking used in?

RCTs

What is the purpose of blinding/masking?

prevent biases stemming from awareness

ensures randomization remains valid

Whaat is single/double/triple blind?

single: blinding one group of people

double: blinding 2 groups of people

triple: blinding 3 groups of people

What is the difference between causality and prediction?

causality: cause & effect

prediction: predicting outcomes

What are 3 criteria for causality/making casual inferences?

1. temporal

2. relationship

3. confounder

Describe temporal criteria for causality

cause must precede effect

Describe relationship criteria for causality

must be demonstrated association between cause & effect

Describe confounder criteria for causality

relationship between cause & effect cannot be explain by third vairable AKA confounder

What is biologic plausibility?

causal relatinship should be consistent w/ evidence frrom basic physiologic studies

What are advantages of RCTs?

most powerful for detecting causal relationships

What are disadvantages of RCTs?

often not feasible or ethical, often expensive

Describe quasi-experimental study design

• no random assignment

• practical

• results may be less conclusive

• C &. E inferences are less compelling than RCT

What is stronger: RCT or quasi?

RCT

Does the researcher have control over treatment quasi-experimental designs?

no

• pre-existing groups are already determined

What are the 3 types of quasi-experimental designs?

1. nonequivalent control group designs

   • experimental group results vs control group results

2. within-subjects designs

   • one group is studied before & after intervention

3. interrupted time series

   • historical comparison

What are advantages to a quasi-experiemental design?

easier & more practical

What is a disadvantage of quasi-experimental designs?

more difficult to infer causality

What is sampling

selection of a portion of population (sample) to represent entire population

What is eligibility criteria?

characteristics that define population

how sample is drawn

What is sampling bias?

systematic overrepresentation or underrepresentation of population

What is strata?

subpopulations of a population; may facilitate subgroup data analysis

• ex. male & female

What is a representative sample?

sample whose key characteristics closely approximate those of population

What are 2 types of sampling designs in quantitative studies?

1. probability sampling

2. non-probability sampling

What is probabiility sampling?

random selection of elements

• each element has equal & independent change of being chosen

What is nonprobablty sampling?

does not involve random selection

• not representative of population

What occurs first: sampling or intervention?

sampling

What are the 3 types of probability sampling?

1. simple random

2. stratified random

3. systematic

Describe simple random sampling

researchers establish sampling frame & then chosen at random

• ex. current MSU nursing students

Describe stratified random sampling

population is divided into 2+ strata & then chosen at random

• ex. divide population into men & women and then choose at random

Describe systematic sampling

involves selection from every kth case

What are the 4 types of non-probability sampling?

1. convenience sampling

2. quota sampling

3. consecutive sampling

4. purposive sampling

Describe convenience sampling

selecting most conveniently available people as participants

Describe quota sampling

identify population strata & determine how many people are needed from each

• ensures diverse subgroups

• ex. ensue male students are represented in nursing study

Describe purposeful sampling

handpicking sample members

• choosing deans for study

Does probability (random) sampling or non-probability (non-random) sampling have more bias?

non-probability

What type of probability sampling has an equal chance of selection?

probability (random) sampling

how are sample size needs estimated prior to study beginning?

power analysis

What is power?

ability of a study to detect a difference/effect when a difference/effect really exists

Does power increase or decrease w/ sample size?

increases

What is effect size?

expresses strength/magnitude of relationships among variables

What are the 3 Belmont principles?

• beneficence

• respect for persons

• justice

What is research misconduct?

fabrication, falsification, or plagiarism in proposing, performing, or reviewing research or in report research results

What is fabrication?

making up data & results

What is falsification?

manipulating data & results inaccurately

What are the 3 criteria for research misconduct?

1. significant departure from accepted practices of the research community

2. ocmmitted intentionally/knowingly

What is authorship?

implies responsibility & accountability for published work

What are the 4 criteria for authorship?

1. substantial contribution

2. drafting the work or revising it

3. final approval

4. agreement to be accountable for all aspects of work

What happens if contributors don’t meet all 4 criteria for authorship?

contributors listed in acknowledgements section