Statistics Flashcards
Symmetrical vs asymmetrical data
Positive skew, to the left of graph
Symmetrical is when graph peaks in the middle and slopes on either side
When to use mean
Symmetrical data
When to use median
Asymmetrical data, outliers in data
When to use mode
Nominal data
Why is range avoided as a statistical aide
Too simple, extreme scores cause problems
What is interquartile range
Middle 50%, from 25 to 75th percentile. Used for skewed data to measure the spread
Problem with interquartile range
Ignores 50% of scores
When is standard deviation and variance used
Shows spread of data around mean, symmetrical data
Relation between standard deviation and variance
Standard deviation = Square root (Variance)
Calculating variance
Sum of (Individual date - Mean) / Number of obs
What are confidence intervals
Ascertaining potential variability around an estimate
What if the in between group difference lies in confidence interval
Result is significant
Is this result significant?
At 60 h, mean reduction in total score was 21.0 points in the brexanolone group compared with 8.8 points in the placebo group (difference -12.2, 95% CI -20.77 to -3.67)
Yes
Is this result significant?
Proportion of patients achieving a global response … was 56.3% with brentuximab vendotin versus 12.5% with physician’s choice, resulting in a between-group difference of 43.8% (95% CI 29.1 – 58.4)
Yes
At 6 months, 285 (47%) of 607 patients and 287 (47%) of 605 controls were dead or dependent (odds ratio 0.98, 95% CI 0.78-1.23). Is this significant?
No as the confidence interval crosses 1.
What values for difference, odds ratio, hazard ratio and risk ratio show no difference
Difference = 0, Odds ratio, Hazard ratio and Risk ratio = 1
Are the following significant?
- Hazard ratio - 2.06 - 0.88 – 4.82
- Difference - 4.58 - 0.85-7.89
- Odds ratio - 0.21 - 0.08-0.55
- Adjusted mean difference - 0.36 - 0.21-0.52
- Odds ratio - 1.20 - 0.59-2.44
- Risk ratio - 1.47 - 0.10-1.98
- Insignificant value for hazard is 1 and CI crosses this
- Significant as CI doesn’t cross 0
- Significant
- Significant
- Insignificant as CI crosses 1, value for odds ratio is 1
- Significant
What is the null hypothesis
The hypothesis that data from a control or placebo is due to change and there’s no difference in the groups. Ex, if testing for a new drug, null hypothesis is that the drug has no effect. We can also come up with an alternative hypothesis, the changes we see are due to the drug having an effect. The null hypothesis holds true unless proven otherwise
What is the p-value
Probability value, probability of observed result given null hypothesis is true. Very low p-value can prove the null hypothesis to be false
For what p-values is the evidence against null hypothesis weak, strong and very strong
p-value < 0.1 = Weak evidence against null hypothesis
p-value < 0.05 = Strong evidence
p-value < 0.01 = Very strong evidence
200 adults seeing an asthma nurse speciailist were randomly assigned to either a new type of bronchodilator or placebo. After 3 months the peak flow rates in the treatment group had increased by a mean of 96 l/min and in the placebo group by 70 l/min.
The statistical test is conducted and a p-value of 0.001 obtained. What does this mean?
The p-value < 0.001 hence the result is significant as there’s strong evidence against the null hypothesis