HSR quiz 2

experiment

operation or procedure carried out under conrolled

levels of evidence

systematic reviews and metaanalysis the best (filtered info)

RCT then cohort studies then case series and reports (unfiltered info)

background info and expert opinion at bottom

purpose of experimental research

study causal relationships

causality established when

IV happens before DV
IV associated with and influences DV

alternative explanations have been eliminated

experimental research and units of analysis

tests hypothesis thorugh planned interventions carried out to make explicit comparisons between or across different intervention conditions

units of analysis may be: individual, family, small group, organizaiton, community 

treatment v control group 

aka experiemntal group, the thing that recieves intervention

• indiviuals or other units that don’t recieve intervention, ideally are as similar as posible to those in experimental

randomization v selection

take group of units and randomly split into experiemental and control.

pure experiment: pull sample from poulation, then randomly assign to treatment or control

bigger samples more representative

randomization is best way to ensure idfferences are due to treatment, not something else 

quasi experiement for hsr/methods to make groups more comprable

matching: nonrandom method to construct control group similar to the experimental group

stratified random sampling- divide people into strata based on charactersitics, then randomly assign 

intercention

IV

takes form of sitmulus, treatment, or program that is present fro the experimental group and is absent for control 

use of placebo

inactive substance that has no therapeutic effect

control for bias since knowing abt treatment can indluence results

strengths and weaknesses of experiements

strong internal validity, relatively accurate inferences about cause and effect

limited generalizability and external validity, feasibility issues, practical challenges, cost and complexity 

retrospective analysis

commonly applied to quantitative data from previous studies or maybe archival qualitative infromation. applies theoretical kknowledge and conceptual skills to existing data to address research q

why valuable for hsr

cost efficient (uses existing data, reducing time and expense compared to primary data collection

large, realworld datasets: population lvl patterns in healthcare use, costs, outcomes

timely and scalable: allows rapid evaluation of policy changes, interventions, system performance

ethically feasible: avoids dierct data collection from patients, minimizing burden and ehtical concerns

generalizable insights: draws on diverse, representative data 

sources of secondary data

administrative data, durvey data, clinical and registry data, policy and program data, proprietary and linked data (EHR)

quanitative methods of secondry analysis

descriptive (distribution, rate, percentage), regression (fixed effects, DiD, regression disocntinuity, synthetic control)

key steps in secondary data analysis

1. define the research question

2. select the appropriate datasets

3. undersatnd data strucutre and documentation

4. rprepare and clean data

5. define variables and measures

6. choose analytic approach 

7. address data limitaitons 

strenghts of secondary data analysis

economy of time, money, personnel

proper documentation

afford opportunity to generate larger samples than primary research

secondary data may be more objective

affors opportunity to study trends over logn erpiod of itme 

weakenesses of secondary data analysis

extent of compatibility btwn data and research question

• must be found rather than created

• searching for data not easy

access of data issues

records incomplete or inaccurate

time consuming 

research positionality

bias in quantitative and qualitative resarch resulting from interaction btwn data and researchers’ backgrounds. care should be taken to avoid reproducing inequality within analytic processes.

qualitative resaerch

undersatnd phenomena on the “why” and “how'“

observations and analyses generally less numerically measurable than quantitaitive.

pros and cons of qualtiative methods in hsr

pro: indepth undersatnding of patient experiences, explore how complex social issues influence heallth, uplift marginalized voices

cons: very time intensive, expensive or requrie a lot f resources, issues abt generalizabiity, feasibility

ethical considerations

IRB: orgnaizations that review and approve research studies involving human subjects

ensure that the rights and welfare of human subjects are protexted

even secondary data may be subject to it

reasons to use qualitative methods in HSR

context for numerical data

reveal underlying social, cultural, environmental determinants of patient engagement or health behaviors

center voices/experitise of the people who are experiencing a specific health outcome

formative research or to gain pilot data

mixed methods 

quantitative purirst view: grounded in positivism, there is one reality, that relaity is objective 

researchers can and should eliminate all biases

qualitative purist view: grounded in constructivism, reality is constructed and multiple realities exist. researcher cna’t be separated from work 

grounded in a philosphy of pragmatism