Research Flashcards
2 types of criterion-related validity
concurrent and predictive
what does validity test?
the strength of the relationships between your outcome (e.g. KAP), what the behaviours are associated with that outcome (e.g. well-informed, have no stigma) and the assessment of those behaviours (KAP test)
what is the 10/90 gap
Idea found in 1900 that 10% of funding is going to diseases that make up 90% of the GBD - so most disease in LMICs were not being funded, even though these cause the greatest deaths
Today, these diseases only account for 45% of the GBD, not 90%
sequential multiple assignment randomized trials (SMART)
adaptive interventions where you change the intervention when you see non-response
TIDieR
Template for Intervention Description and Replication
Checklist to improve the completeness of reporting, and ultimately the replicability, of interventions because papers were being too vague and useless.
content validity
is it comprehensive? does it cover all the things associated with that property?
e.g. a KAP test without any questions on knowledge would have low content validity
sensitivity
the ability to corrently identify patients with a disease
p-value
The probability of seeing the observed result assuming the null hypothesis is correct.  
The p value is the evidence against a null hypothesis. The smaller the p-value, the stronger the evidence that you should reject the null hypothesis.
summative evaluation
aims to understand the impact of the intervention and its implementation at the end of the study. Whilst formative evaluations seek to improve implementation, xx evaluations seek to look more broadly, often including economic evaluations
concurrent validity
is there a relationship between your test and another previously validated test?
e.g. your new measurement for intelligence and an IQ test show the same results = strong concurrent validity
epilepsy is a neurological condition that affects the _____
nervous system.
predictive validity
can our test predict future outcomes?
e.g. does your KAP test allow us to make predictions about how supportive you will be towards PWE
Which is largest? Pubmed, Medline or Embase
Embase
Then PubMed, then Medline
construct validity
does this (instrument or experiment) measure what it is supposed to measure?
does this construct represent real-life situations?
scoping review
a type of evidence synthesis you do before a systematic review to determine how much research is out there and to find knowledge gaps 
scoping reviews help decide if you should do a systematic review and what the specific systematic review question should be
Which databases include Medline?
Embase and Pubmed
what types of seizures are there?
focal, generalised or unknown onset
face validity
does it look like it measures the property at the surface level? It appears to be related. 
e.g. KAP test looks like it has the appearance that would tell us something about KAP
criterion-related validity
is the assessment related to some other criteria? Is there a relationship with other outcomes?
e.g. is there a relationship between KAP test results and behaving like someone who doesn’t stigmatise.
which is more structured, a literature review or a scoping review?
scoping review
validity
are we assessing what we want to assess
5 main constructs of CFIR
intervention characteristics, 
outer setting, 
inner setting, 
character of individuals and 
process
telemedicine
caring for patients remotely via electronic devices.
definition of epilepsy
at least 2 unprovoked seizures occuring more than 24 hours apart
formative evaluation
like a process evaluation but with feedback and the aim to change and improve during the study
confounder
associated with the independent and dependent (outcome) variable but not on the casual pathway. 
an extra variable that you didn’t account for, creating a hidden effect.¬†
e.g. lack of exercise –> weight gain
how much people eat could be a confounder as that would also add to weight gain
selection bias
people in your study differ from the population of interest i.e. not representative
three most common types of quasi-experimental designs
1. uncontrolled before and after studies
2. time series designs
3. controlled before and after studies
social desireability bias
subjects choose responses they think researchers want to hear or know are the ‘right’ answer
specificity
the ability to correctly identify people without the disease
spurious
not genuine, not true
model
simplified depiction of a more complex world
convulsive seizures are also known as ___ and involve ___
generalised tonic-clonic seizures
whole body shaking movements and loss of consciousness
impedance
the amount of opposition to an electrical current i.e. how much it impedes, hinders, delays the electrical current in a circuit. High impendence will mean poor signal quality
seizures may be casued by…
brain injury, genetics, brain structure, etc.
process evaluation
observational research on how an intervention is being implemented. No feedback for how to improve.
test-retest reliability
a way to see if your survey is consisten over time. 
get people to take your survey, then when a while, then get them to do it again - compare the scores.
moderator / effect modifier
the size of the effect on A –> B
e.g. length of time on instagram –> increased social comparison but the effect of the social comparison is moderated by your self-esteem. If you have high self-esteem, the effect of social comparison will be less.
noramalisation process theory
theory of how innovation is accepted - study of culture
mediator
causal relationship where A –> M –> B, so the xx can produce changes to the outcome variable.¬†
e.g. being hungry –> passive aggressive (M) –> argument with friend
being hungry doesn’t lead to arguments with your friend, your passive aggresive mood does.
routine data
data collected at regular intervals
what are the two components that make up the h-index
quantity of papers published
quality of papers published (measured by number of citations)
pseudoanonymised data
potentially identifiable as data may be attributed to an individual if linkage information can be accessed elsewhere by researchers.
e.g. replacing names with numbers but being able to access names on a different document
within subject comparison
Subjects are assessed before and after an
intervention and results analysed in terms of within subject changes
All participants get every treatment.
thematic analysis
v common type of qualitative analysis where you identify patterns (‘themes’) in the data
the NASSS framework
Trisha Greenhalgh’s work
non-adoption, abandonment, scale-up, spread and sustainability 
framework to assess the failures of tech implementation
seizure
abnormal neural activity in the brain resulting in uncontrollowed shaking with loss of consciousness (tonic-clonic) to subtle momentary loss of awareness (absence seizures).
name all the implementation outcomes
acceptability, adoption, appropriateness, costs, feasibility, fidelity, sustainability
sampling bias
some people are more likely to be selected in the sample than others 
what are the two things that cause statistical heterogeniety in systematic reviews/meta-analyses?
clinical diversity 
methodological diversity
what does it mean if the dymond in a forest plot is on the line
there is uncertainty, it is not clearly on one side 
