Please enable JavaScript.
Coggle requires JavaScript to display documents.
GENERAL COMMON THEMES/ISSUES FOR CRITICAL APPRAISAL, NEED TO KNOW WHAT…
GENERAL COMMON THEMES/ISSUES FOR CRITICAL APPRAISAL
SAMPLING
Data Saturation
Small Sample Sizes
which are not justified with any sort of power calculation
Sampling Methods
Convenience Samples
furthermore many of the studies the students who volunteered are likely to be the ones doing well.
Purposive Sampling
of participants
of hospital locations as well ( chance of bias)
tend to only be from one university
davies et al example
Snowball sampling
Demographics
Age
tends to be more women than men in studies ( though some counter argue that this is proportionate to the physio workforce anyway)
many studyies disproportionate do not mention older students ( they are outliers)
Geographical Location
Many studies are only done in one university ( see sampling metods comment above)
A large proportion of the good papers from Australia
not as gneeralsable as different geographical locations and services differ in what they provide.
different placement structures etc
Disproportionately metropolitan ( though to be expected as most people live in cities anyway)
some done in America
some different themes emerged from this like getting sued??? so should have been more rigid with exclusion criteria
their students tend to be older
some studies provided better demographic info and others very little ( find examples for assessment)
Disproportionate Samples
Gender
many more women than men in most studies ( though one paper counter pointed that this is proportionate and reflective of the physio workforce
age
disproportionately young subjects
older people were outliers
STUDY METHODOLOGIES
Bad
Lots of Cross sectional studies
lack of longitudinal studies
not really any systematic reviews specific to physio of any use ( there is one for nursing )
Lack/ no control group for some studies (not really many RCTs
some of the pilots were bad ( one example the students carried them out)
some instances of potential bias in papers
As lots of the good research is not being done in UK they are only generalisable to a certain extent.
differences in services etc
Heterogenity in study design
different papers quantified new grads as different things
some they only had to be a year others they could be up to 5 years
Differences in interviewing/collection methods
Duration
some very long ( 2 to 2.5 hours in Atkinson and McElroy)
some very short
Type
Most semi-structured using focus groups
some telephone
some surveys
differing times to look at and reflect on the questions
some studies were not as reflexive as you'd want ( didnt keep logs of their thinking)
many studies only one uni or hospital
some conducted by lesser experienced reserachers (Mc elroy example)
include pilot studies here
some studies no transcripts ( Merga example)
Good
many of the studies did pilot studies
independant researchers to reduce bias
most had an ethical considerations section and asked their participants for consent
Qualitative Specific
many studies provided transcripts of what they asked their participants
many used validated coding software for producing themes
some papers mention data saturation for their themes
most conducted semi- structured interviews
Pros
cons
in some instances Participants given the chance to fully prepare and think of what all is important to say - can form their opinions/answers outwith the pressure of the focus group and then present their talking points
In many of the papers the researchers read the papers multiple times to whihc could ensure their analysis of contect is not misinterpreting what the participants are saying
multiple researchers to triangulate
many claimed they had reached a data saturation [point
By and large inclusion criteria for most of thestudies was specific engouh ( though of course some not specific enough)
Generally aims clear and well defined by most papers
most papers had funders but these were largely universities etc
some studies used variety / multiple hospitals to triangulate data = good
some papers had very good practical applicability in their implications sections( Merga example)
response rates generally alright in most cases ( b this might in part be due to low sample sizes)
SPSS tool generally used for quantitative researc ( gold standard ( davies et al example)
RESULTS AND ANALYSES
STUDIES THEMSELVES
many of them are older ( aroudn 2010 ish)
Some of them are very niche and therefore hard to generalise?
this coupled with small sample sizes (see above) means alot of paper stretched thier inferences and jamp to conclusions.
equally though, barradell et al was too broad and tried to do too much therefore the inferences they were making were likely to be innacurate
some studies there is a large gap between the time their data is collected and the actual date of publish which makes the reader question whether what they have written is relevant
davies et al example
some papers their implications sections are too bogged down by theory and therefre are unlikely to make clinical impact
limited number of quantitative studies though this is most likely due to our search strategy
NEED TO KNOW WHAT GENERAL DIRECTION WILL GO DOWN FIRST FOR EVERYTHING ( DELETE ALL SAMPLING BITS)