Term 1 (Chris) Flashcards
What is PICOT? What does it stand for?
A framework for evidence-based research questions
P - Patients / population - who will be participating
I - Intervention/interest (exposure) - what is being tested
C - Comparison - what is the comparison group
O - Outcome - what is the outcome or endpoint
T -Time - when should outcome be measured
What are the two types of primary research designs
Qualitative
Quantitative
what are the three types of qualitative designs
interview
focus group
observational
what are the three types of quantitive studies
survey (closed)
observational analytic
experimental
what are the 3 subdivisions of observational analytic
cohort study
case-control study
cross sectional study
what are the two subdivisions of experimental studies
randomised control trials
non-randomised
what are secondary research designs
systematic review or meta analysis
what are the strengths of randomised controlled trials
provides evidence of causality
rigorous evaluation of single variable
what are the limitations of randomised controlled trials
resource intensive: costs time and money
needs a large number of participants - many studies underpowered
ethical challenges
what is the difference between a prospective and retrospective cohort study
In a retrospective cohort study, the group of interest already has the disease/outcome. In a prospective cohort study, the group does not have the disease/outcome,
what are the strengths of cohort studies
can establish population-based incidence
can study several outcomes for each exposure
can establish cause effect
what are the limitations of cohort studies
resource intensive: costs money and the
large number of people needed
loss to follow up
inefficient for rare conditions
what are the strengths of a case control study
small sample size needed
appropriate or studying rare conditions or those with long lag between exposure and outcome
what are the limitations of a case control study
exposure assessed after disease occurrence
reliance on records to determine exposure status
highly susceptible to selection bias
what is the purpose of cross sectional studies
document health status in specific population at a specific point
what are the strengths of cross sectional studies
provides estimates of prevalence of a disease
can identify population healthcare needs
easy fast and inexpensive
no follow up required
what are the limitations of cross sectional studies
cannot determine casual relationships
participants may provide social desirable answers
impractical for studying rare diseases
why would you use a qualitative study
useful for understanding patients experiences perspectives and views
what are the strengths and limitations of qualitative studies
strengths:
enables understanding of patients experiences, unpredictable and insightful findings
limitations:
difficult to generalise, sample selection based on certain experiences
small sample size
what are meta analysis / systematic reviews used for
answer a specific clinical question
combines results of previous studies to produce one overall measure of the effect of an intervention
what is the difference between descriptive statistics and inferential statistics
descriptive - techniques we use to describe the main features of a data
inferential - statistical inference is the process of using the value of a sample statistic to make an informal guess about the value of the population parameter
what is the difference between categorical and numeral data types
categorical - sex, blood type
number - age, weight, number of siblings
what are the two types of categorical data
nominal and ordinal
what are the two types of numerical data
continuous and discrete
what is the difference between nominal and ordinal
nominal - no natural ordering ie sex or blood type
ordinal are ordered in severity or disease stage
what is the difference between continuous and discrete data
continuous - no value limitation such as weight
discrete - whole values only such as hospital visits
what are some ways you can present categorical data
bar chart
pie chart
frequency distribution
what are ways you can present data that is numerical
histogram
box or whisker plot
what does a frequency distribution table show
the frequencies and percentages in each group or category
what do these plots/charts show:
bar chart/pie chart
scatterplot
box plots
box - display frequencies of categories (categorical data)
scatter - numerica data using two continuous variables
box plots - summary statistics for numeric data
what is positive / negative skew
Positive Skewness means when the tail on the right side of the distribution is longer or fatter. The mean and median will be greater than the mode
Negative Skewness is when the tail of the left side of the distribution is longer or fatter than the tail on the right side. The mean and median will be less than the mode.
what is the difference between the mean median and mode
mean = average
median - middle number
mode - most common number
what does the mean median and mode not tell us
the spread and range of data
what does the range do
indicates the extremes - min and max value
what is the standard deviation
average of values around a mean
the larger the SD the further away the values re from the mean ie greater spread of data
how do you calculate standard deviation
Calculate the mean Subtract the mean from every value Square these new values, and add up Divide this total by (n-1) = variance Take the square root = sd
what does the interquartile range do
summaries the spread of values around the mean
its a range from the lower quartile (25%) to upper quartile (75%0
what values do you report for normally distributed data compared to discrete data
normal: report mean and standard deviation
skewed data: report median and interquartile range
what values are not dependant on the distribution
mode and range
what are the two types and sub types of research
observational - cohort and cross sectional survey
experimental - lab experiments and randomised clinical trials
what are the two types of statistical inference
estimation - using summary statistics from collected sample to represent the population
hypothesis - making hypothesis and investigating whether there’s evidence against the hypothesis
what is a confidence interval
gives us a plausible range of values (96% of the times) if we were to repeat the experiment
what does correlation measure
strength of a linear relationship
what are the different types of correlation
Ifr> 0 we have a positive correlation; implying that as one variable increases then so does the other.
Ifr< 0 we have a negative correlation; implying that as one variable increases then the other decreases.
Ifr= 0 we have no correlation; implying there is no association between the two variables.
If r=+1 there is a perfect positive correlation.
If r=-1 there is a perfect negative correlation.
which correlation due you sue is data is not normally distributed
use spearman rank as it is less sensitive to strong outliers
when do you not use correlation
when data is not liner or there are distinct sub groups
which tests would you use for one numerical and one categorical association
indépendant staples t test - two sample t test
Mann-Whitney-U test
what are the limits in a two-sample t test
-2 - +2
what are the three assumptions of two sampled t tests
two independent groups
numerical variable is normally distributed
similar standard deviations
how would you examine an association between two categorical values such as survival and treatment
proportional difference
what are the three types of proportional differences for measuring two categorical associates
chi squared - standard
chi squared - yates correction
fishers exact test
what are the assumptions for chi squared tests
two independent groups
number of expected values in each of four cells should be greater than 1
in three of four cells the expected value should be greater than 5
what is the problem with using yates correction
For small sample sizes the chi-squared test is too likely to reject the null hypothesis. A continuity correction can be made to allow for this.
what is wrong with using fishers exact test
If a contingency table fails to meet the conditions required for the chi-squared test then Fisher’s exact test can be used.
what do inferential analyses do
draw conclusions about populations from staples of data drawn from the population
what does a sample do
measurements are used to estimate population parameters
what is standard error
represents the average distance between an estimate and its population parameter
how do you calculate standard error
Calculate the mean (of all the samples’ means)
Subtract the overall mean from every sample’s mean
Square these new values, and add up
Divide this total by (n-1) = sampling variance
Take the square root = standard error (= standard deviation of the sampling distribution of the mean)
how do you calculate the standard error of man and what does it tell us
SE = standard deviation / square root of number of samples
it will always be smaller than the standard deviation
the larger the sample size the smaller the standard error
how does SE show us 95% confidence intervals
estimates the range that is most likely to occur - 95% of all sample means will fall between our sample mean value
how would you calculate 95% confidence for the upper and lower interval
Using our example of n=10 GP surgeries, mean=4, SEM=0.79
The 95% confidence interval for the GP surgeries data is
4 - (1.96 x 0.79) = 4 -1.58 = 2.42 (Lower limit)
4 + (1.96 x 0.79) = 4 + 1.58 = 5.58 (Upper limit)
The interval from 2.42 to 5.58 is therefore called the 95% confidence interval (95% CI) for the mean
what is null hypothesis
there is no difference between the two groups
what does a p value do
significance level 0.05 - allows us to reject or accept null hypothesis
what is the difference between a large and small p value
large p value: quite likely to see results by chance
cannot be sure of a difference in the target population
small p value:
unlikely to see results by chance
there may be a difference in the target population
cuts off point is arbitrary
if very unlikely then reject the null hypothesis
what does p value greater than 0.05 mean
no statistical difference - accept the null hypothesis
what is the relationship between p value and 95% CI
If 95% CI of a mean difference does not contain zero, reject the null hypothesis at statistical significance (p-value of 0.05) of 5%
what are the different uses for p value and confidence interval
p value: weight of evidence to reject null hypothesis no clinical interpretation confidence interval: can be used in hypothesis testing clinical interpretation
what is cares
clinical assessment reasoning ethics safety
what is qualitative research
collection and analysis of non-numerical information
why and how of population behaviours experiences etc
what are the general differences between qualitative and quantitative data
quant - findings are generalisable
qualt - unique and insightful
what are the types of qualitative research methods
Content Analysis Grounded Theory Framework Analysis Protocol Analysis Ethnography Phenomenology Discourse Analysis Conversation Analysis Thematic analysis Interpretative Phenomenological Analysis (IPA)
what is a theme
themes are recurrent and distinctive features of participants accounts which the researcher sees as relevant to the research question
it can reflect a pattern of response which is determined by the researchers judgement
what is the value of qualitative studies
enables understanding of patients choices and why they did this
what is AnSWeR
antenatal screening web resource
What are the limitations of findings in qualitative research
generalisability
sample selection based on certain experiences (not random)
sample size often small
reliability - findings based on interpretations by the researcher
what are the strengths of qualitative findings
robust methodologies
offers in depth understanding of diverse perspectives
unpredictable and insightful
usually unobtainable using a quantitative approaches based on preconceived ideas
what is a systematic review
uses systematic and explicit methods to identify, select and critically appraise relevant evidence
involves collection and analysis of data from the studies that are included in the review
what is meta-analysis
stat technique which summarises the results of several studies into a single estimate
what is the relationship between systematic reviews and meta analysis
systematic reviews often but not always include meta analysis
why do we have systematic reviews
information management reduction of subjectivity / bias increase power and precision identify gaps in research efficient use of resources
what are the criteria for Mulrows review quality
purpose data identification study selection criteria validity assessment qualitative/quantitative synthesis summary future directives
what is cumulative meta analysis
repeated performance of meta analysis whence a new relevant trail becomes available for inclusion
what are the key stages in a systematic review
define the question
look for all studies reliably addressing the question
sift the studies to select relevant ones
assess the quality fo the studies
calculate results for each study
interpret results
what specifically makes a systematic review
was there a clear research question
was protocol developed
are the inclusion criteria clearly stated
what is heterogeneity
between study differences: some random difference between studies is expected due to random variation
what are the there types of heterogeneity
Clinical: different patient populations, interventions, follow-up times, choice and measurement of outcomes.
Methodological: different study designs, quality issues (e.g. not all trials described as RCTs may have adequately randomised patients)
Statistical: numerical variation in treatment effects
what is publication bias
the selective bias of research evidence based upon the results of that research
positive studies more likely to be published than negative ones
english language, availability
what are the four conclusion of systematic review
we now know for sure that it works
we now know for sure it doesn’t work is no better than..
the research is of such poor quality or is so biased that
there is no research
