Topic 1 - The Scientific Approach Flashcards
more information doesn’t always mean…
more answers. Also lots of misinformation
define the difference between misinformation and disinformation
Misinformation: passing on wrong info, usually accidentally
Disinformation: purposely spreading wrong info, purposely deceiving people
Name 7 tips on stopping the spread of fake news
1) Access the source
2) Go beyond headlines
3) Identify the author
4) Check the date
5) Examine the supporting evidence
6) Check your biases
7) Turn to fact-checkers
Science is never_
black and white
- always evolving, science is never 100%
the goal of stats isn’t to confirm/deny something, it’s to_
look at the probability of things
A lot of misinformation doesn’t come from nothing, give some examples shown in class
- hydroxychloroquine: used to treat malaria and symptoms of some chronic diseases, gained traction for potential covid treatment
- ivermectin: used to treat parasitic worms in animals, gained traction for potential covid cure
*both had promising results very early
Although there is no “scientific proof” and concrete answers rarely exist, _
research provides “evidence
*heathy scepticism is what science is built upon!
The term scientific literacy is _
super broad
Define scientific literacy
- a skillset and mindset that will support problem solving and making intelligent, informed decisions
- a way of thinking that is more logical, with a healthy level of skepticism
define intuition
- draw general conclusions based on emotions and instincts
- can be based on past experiences/partial evidence
*always gonna have some level of intuition, but have to move past it to the other ways of knnowing
define authority
- statements from authorities must be true
- persuation does not equal to credibility
- coming from people who are experts on topic/ppl with more power and knowledge in field
define rationalism
- reason/logic to draw conclusions (without actually seeing anything)
- what if our initial premise is wrong…
- if_is that, then_is_
*if initial premise is incorrect, our logic becomes faulty (ex. loss of smell and covid status)
define empiricism
- making conclusions through structured observations
- different than anecdotal evidence
**the scientific approach is grounded in empiricism!
the scientific approach is grounded in empiricism BUT _
also relies on rationalism, authority and intuition
*can’t always rely on empiricism
*healthy scepticism, don’t blindly accept what you hear and read
*evaluated on the basis of logic and the quality of scientific results (aka. not the existence of results)
define the scientific approach
systematic set of principles and procedures for generating knowledge
- science is very incremental, but usually very cyclical approach: adding more info which generated new question
list the 7 characteristics of the pseudo-scientific approach
- hypotheses not testable
- methods are not scientific or validity of data is questionable
- evidence anecdotal
- heavy focus on “experts”, not scientific references
- ignore conflicting evidence
- use many “scientific-sounding” terms/ideas
- claims are vague
what are 2 types of common misinterpretations of scientific data?
statistical misinterpretations, overgeneralizations
name the 3 components of statistical misinterpretations
1) correlation does not equal causation
2) confounding variables
3) statistical significance does not equal importance
name the 2 components of overgeneralizations
1) a single study is rarely confirmatory
2) results may apply to a specific group
define correlation
a relationship between 2 measures
- reported as correlation coefficient (0=not related, 1=perfectly related)
describe correlation does not equal causation
- just because 2 things are related to one another, doesn’t mean they cause the other
ex. taller/bigger shoes, eat nuts/live longer
define confounding variables
other variables that may also influence the relationship but are not the primary focus
*when these are ignored, might make it seem like a relationship is more meaningful than it is
define statistical significance
- describes how likely what we observed is due to chance
- our confidence in our results
*p-values (smaller = less likely to be chance, 0.05 threshold in most research but debated)
*influenced by average scores, how much they vary, and # of scores collected
Statistical significance is more than just about means, how could the average be same but significance be different?
the spread of data values could be larger (variability of scores/observations)
*smaller spread is better because less difference between people = confidence in results
OR more people in study
statistical significance does not…
equal importance
why is a single study is rarely confirmatory?
- each individual studies provides incremental evidence
- best to examine the entire body of evidence
*overgeneralizing
why is it that results may apply to a specific group? overgeneralizing*
- studies are usually very specific in their recommendations (don’t extrapolate results to other groups easilty)
- media is usually not as specific in their interpretation
ex. walking faster may lead to longer life
what does it mean for research to be incremental?
- build upon previous finding
- adding evidence, goal is to build on existing findings a lot of the time
how is the ‘levels of evidence’ period organized?
lowest quality of evidence at the bottom, highest at the top
name the 5 levels of evidence in order of lowest quality of evidence to highest
evidence from expert opinions, case studies/cross-sectional, cohort studies, randomized controlled trials, reviews/meta-analyses
describe cohort studies
- group of people observed over time
- need people to be free of the disease/outcome at onset but differ in the exposure (yes/no)
*ex. can’t run experimental trail on smoking, have to use cohort
name the advantages for cohort studies
- establish a temporal relationship between exposure (smoking) and outcome (cancer)
- evidence towards cause and effect (not 100%)
name the limitations for cohort studies
- expensive/time consuming
- difficult for rare disorders
- can be “confounded” with additional factors that interfere with/influence outcome
(ex. were smokers vs non smokers different in other ways? other activities that expose them to greater risk?)
describe randomized controlled trials
- experimental group (treatment) + control (no treatment)
- random assignment is necessary
- placebo is important BUT NOT NEEDED (hard in surgery/interventions)
list the advantages of randomized controlled trials
- highly controlled
- best design for cause and effect
- randomization eliminates/minimizes confounding variables (unless there are 3/4 of 1 type in one group, usually look at stats to see if even)
list the limitations of randomized controlled trials
- not always possible/practical
- ethical considerations
- need people to follow their assignment
in the RCT knee surgery ecample, how did outomes (self-reported pain and function) compare 2 years after the surgery?
same at 2 years for sham and treatment
describe systematic reviews/metaanalyses
- search out all studies on a given topic to summarize the findings
- systematic reviews can be qualitative/quantitative
- meta-analysis: quantitative pooling of results
