Ress exam Flashcards
What is the epidemological triad?
Place, time person
What does incidence meam?
The number of new cases in a given time
What is the incidence rate?
The number of people who get a the diseaes in pre defined time/ the number of people at risk (who can get the disease)
What is prevelance?
The number of people who a disease amongst a population at a specific time.
what is the effect of short duration and long duration illness on prevelance and incident?
Short duration illness will have a high incident but a short prevelance
Long duration illness will have a low incident but a high prevelance
Prevelance good for chronic disease
What is case fertality rate?
The number of people who die from a disease in a given period/ the number of people with the disease
What is mortality rate?
The number of people wh die due a diseae in a period/ the number of people who die in a period
what is the aim of “adjustment”?
To reduce the effect of a factor that may swerve the results.
For example adjustment to compare people of a high average age in a population and a low average age in a population
How is “adjustment” done?
By statum specific rate e.g. sex and age
What is risk?
The probability of a disease occuring
1–> certain to happen
0–> Will not happen
How do you calculate the risk?
The number of new cases/the number of people at risk
Give a example of risk?
You apply risk to a group not a person as a person is either going to get the disease or is at risk of getting the disease.
For example 1 in 5 people will get lung cancer who smoke for example.
Therefore there is a 0.2 risk of getting lung cancer if you smoke
What is the aim of the risk ratio and in what study will risk ratio be used in?
Risk ratio aims to find out the risk factor (exposure) of the risk–> the relative risk
It is only used in cohort studies
What do the relative risk values mean?
Less than 1, more than 1 and = 1
If the relative risk = 1 then the effect of exposure is same as unexposure
If the relative risk >1 then the exposure associated with harm
If the relative risk is
How do you calculate the relative risk?
The risk for the treated group/the risk of the control group times by 100
What is relative risk reduction?
100- relative risk.
The reduction in risk in the treated group compared to the untreated group
What is odds?
The probabilty that a event does occur/the probablity a event does not occur
What is the difference in use between risk ratio and odds ratio?
Risk ratio only used in cohort studies while odds ratio is used in case/randomised control studies
What studies is odds ratio used in?
Case/randomised control studies
What is dichomotous variable?
When it can only be one of 2 variables?
Dead or alive
DIsease or undisease
what is population?
Every person in a defined group of interest
What is categorical variable?
Results that can be assigned to a single category
For example male female
Black, blue, red
What is the subdivision of Categorical variables and give examples?
Nominal–> Category with no natural ordering. E.g. Male and female. Black, white, yellow etc
Orderinal –>Category with natural ordering. For example the severity of pain. Very bad, bad, okay , good.
What is numerical values?
Ones that take numerical values
What are the two subdivisions of numerical values?
Discrete –> whole numbers e.g.Number of hospital visits 81, 72
Continous–> no restriction on value e.g. Weight can be 81.245kg , 74,34kg
What is frequency?
The rate of somethign happening
What does the frequency table show?
The frequency and relative frequency
What type of data is bar charts appropiate for?
Categorical or discrete numerical data
What type of data is pie chart used for?
Categorical or discrete numerical data
What is the difference between histograms and bar charts?
Each bar in bar chart is seperate
While in histogram it is continous
When histograms used?
For the display of frequency and continous variables
What is the mean?
The average of all the values
What is the median?
The middle value of all the values
What are the different quartile ranges?
Q1: lower quartile is the median of the lower values. 25% of the values are below this
Q2: the median. 50% of the value are below this
Q3: Upper quartile, the median of the upper values, 75% of the value are below this.
What is standard deviation?
What do you square root
It is only appropiate for numerical value.
It is the measure of the spread of the values from the mean.
The greater the standard deviation the greater the spread of the values from the mean.
The square root of the variance
What does it mean by outcome and exposure?
Outcome is what you are measuring
While exposure is what affects the outcome
If the outcome is numerical and the exposure is numerical what is the analysis?
Correlation
If the outcome is numerical and the exposure is categorical (independent groups) what is the analysis?
T test or Mann Whitny U test
If the outcome is numerical and the exposure is categorical (non-independent groups) what is the analysis?
Paired T test or Sign test
The groups are related to one another
What is the confidence interval?
Usually at 95% where if the results were done 100 times then 95 times the results will be within that range
What is standard error?
The measure of how reliable the sample population measure is of that population.
The measure of precision
What is standard error used for?
It is used to calculate the confidence interval
How do you measure the standard error?
The standard deviation/root of the sample size
What three things impact the width of a confidence interval
Confidence level
Variability–> measured by standard deviation –> more variable populations have wider confidence interval
Sample size –> the smaller the sample the wider the confidence interval
How does a small and large sample effect the confidence interval?
When there is a small sample (200) there is a narrower confidence interval. Meaning the confidence interval is between two numbers that are close.
Use a known SD and normal distribution tables using z-values. (z=1.96 or -1.96)
What is the P value?
The probability that something is happening
What can be said if the P value equal to or less than 0.05?
Then this indicates that the results have occured not due to chance. It is statistically significant.
What is the inductive hypothesis?
Hypothesis is done after the analysis. You know there is a relationship and want to test for it.
You go from being specific to more generalised.
From the bottom up
What is the deductive hypothesis?
Hypothesis made befre the analysis. You are not aware that there is a relationship but want to find out if there is.
Want to see if there is a relationship between two variables
Go from being more generalised to specific
Top down aproach
What follows from doing a scientific method then obeservation?
You make a hypothesis
What is another name for point estimate and where does the values lie on either side of the median?
Termed the standard error and the half the values are on either side of the median
What dos the null hypothesis (Ho) usually conclude?
That something for example a drug has no effect
What is alternative hypothesis (H1)?
That a drug has a effect, but the effect can be harmful or beneficial
If the T test value is greater than the confidence interval should you accept or reect the null hpothesis?
Reject the null hypothesis
If the T test value is lower than the confidence interval should you accept or reect the null hpothesis?
Accept the null hypothesis
What is association?
Comparing two variables
What is correlation?
When we have continous data between two variables and a linear regression line can be created
What is Pearsons correlation r?
If we have a linear regression then we use the pearson correlation (r) to compare the linear correlation between two numerical values. It is a measure of how well the data fit the straight line
The R value is between -1 and 1
What does it mean if the pearson correlation value is:
>0
R>0 then there is a positive correlation between the two variables. As one increases so doe the other
R
Below what pearson correlation should you be suspecious of the correlation?
Correlations of less than 0.7 should be treated with suspicion.
When do you not use correlation?
When there is no linear relationship between the two variables
There is outliers
There is a definitive subgroup
When one or both of the variables is not distrubuted correctly
When one or both of the variables are not numeric
What does sensitivity of a test mean?
How well does a test identify a condition.
Number that correctly test positive/ number of people with the diseae
I.e does it identify everyone with the disease
What does specificity of a test mean?
How well does it correctly exclude people who does not have the condition.
Number of people identified not having the disease/ the number that don’t have the disease
I.e How well does the test identify only those that do have the disease
What is positive predictive value?
The probability that someone has the condition if they test positive.
The number that have the disease/the number that test positive
What is negative predictive value?
The probability that someone doesnt have the condition if they test negative
The number of people without the condition/the number that test negative
What is validitiy?
When you accurately measure what you are meant to measure.
How well does the test actually measure what it is meant to measure
What is reliability?
Test producing the same result if repeated .
Test produce stable and consistant results
What is responsivness?
How well can a test detect real change as they occur
What is primary prevention?
Give examples?
Preventing a future occurence occuring in a unaffected individual by removing the cause.
Aim to prevent the disease before it occurs
Interventions include vaccination programme, education about health
what is secondary prevention?
Give examples
Preventing clinical disease by screening, early detection and/or treatment
Aim to reduce the impact of the disease or injury that has occured.
Widespread screening programmes and taking aspirin
What is tertiary prevention?
Give examples
Prevention of the disease by treating the clinical cases.
The aims to soften the impact of an ongoing illness or injury that has lasting effects.
For example stroke clinics and long term physio
Screening is a diagnostic tool that is a secondary prevention. What factors does screening need to be effective?
To be effective the screening has to be a applied to a wide population
Be quick and inexpensive
What is a gold standard test?
Is a diagnostic test in regards to as defenitives to whether a patient has a diseae or not.
What is the effect to the t distribution as there is a greater degree of freedom? The degree of freedom goes to infinity
The T distribution become the same as the normal distribution
Do we need to consider the clinical significance of a study where the results are not statistically significant?
Yes –> It is possible to get a non-significant statistical result when there is a real clinical difference.
What is the step by step of a research process?
1) You define your questions
2) gather information
3) form hypothoesis and design method ( where you do ethical approval)
4) collect data
5) analyse data
6) interpretate data and form conclusion
7) discuss findings
8) commiunicate findings
When do you not reference?
When it is considered common knowledge
When its your own opinion
When its your own research/data
What should a last name in the reference be finished with and how many authors can be named before yu use ‘et al’
You finished with a full stop
After 6 or 7 authors
What is simple random sample?
The only true random sample. Where the population chosen are truely random
What is stratisfied sample?
Samples are randomly chosen but set into groups to ensure that different groups within the population are representated depending on there size relative to the population.
For example in NHS greater number of nurses will be selected than doctors as there are mre of them
What is quota sample?
You are told what sample to take for example males aged between 35 to 45yrs.
This is subject to interview bias
What is interval sample?
You take samples at set intervals.
Such as a drug company testing the every 100th drug
What is cluster sample?
Where you divide populations into clusters such as north yorkshire or Devon
Give example of continous data?
Height and weight
Give example of discrete data?
Number of rooms
Number of workers
What calculations are done on measure of location?
Mean, median and mode
What calculations are done on measure of spread?
Range, SD and interquatile ranges
What is the indication of positive skewed data?
The data is skewed to the right and therefore mean>median>mode.
Also means that the upper quartile range has more values than the lower quartile range. Show a greater spread
What is the consequence of a negative skewed data?
The complete opposite.
Mode>median>mean
The data is skewed to the left and the lower quartile range will show a greater spread.
Give examples of Boolian connectors/operators?
Words such as and ,or,not that help to narrow down the search
What would a ASTERISK do?
Would allow a truncated ending, where alternative endings are allowed
What do Scholar journals assume?
That the reader has prior knowledge of the subject
What are the 3 categories of journals?
Primary: The research done by a person in various capacities
Secondary: Summary and conclusions made of primary research
Other
What is a cohort study?
Follow a population with a certain trait over a period of time
What is case control study?
You compare two groups of people.
Group with the disease and group without the disease (control group)
You are comparing the disease characteristic in a control group and disease group
Waht is pervelance study?
What is the anme if the study is done over a long period of time?
To identify the prevelance of a disease and if done over a long period of time called a longitudinal study.
Type of research survery.
What is metanalysis?
It is integration analyis of data from several studies
What is relative frequency?
The rate of something occuring/ all the outcomes that occur
What does it mean when there is a pearson correlation of 1?
There is a perfect positive correlation between the two variables
What does it mean when there is a pearson correlation of -1?
There is a perfect negative correlation between the two variables
What does it mean when there is a pearson correlation of 0?
There is no correlation at all between the two variables?
They do not affect each other
What does it mean when there is a pearson correlation is less than 0?
That as one variable decreases the other variable increases
