BIOSTATS Flashcards
What is ordinal data?
Observations that can be ordered/ranked according to some criteria. e.g. on a scale such as 1-10 or asking someone if there pain is high,medium or low.
What is nominal data?
When observations are classified into separate categories that have no logical ranking e.g blood types AB, O .
Is the data nominal or ordinal if there are only two categories? (binary data)
Always nominal
Smoking status of visitors of trauma patients at The Alfred (strict no-smoking policy) in 2010. (NOMINAL, ORDINAL, CONTINUOUS OR DISCRETE??)
Categorical, nominal
The standard categories of Body Mass Index: < 18.5 (underweight), 18.5 – 25 (normal), 25 – 30 (overweight) and ≥ 30 (obese). (NOMINAL ORDINAL CONTINUOUS OR DISCRETE???)
Categorical, ordinal
ody Mass Index, calculated using patient weights and heights. (NUMERICAL, ORDINAL NOMINAL OR DISCRETE??)
Numerical, continuous
Can the sample size be bigger than the population size?
YES. it can bigger than the population size if redundant samples are taken
When is SRS (Simple Random Sampling) used?
Simple experiments tat require a single sample to be taken from a population , target group is large,
When is stratified sampling used?
Used it when there are smaller sub-groups that are to be investigated, you want to achieve greater statistical significance in a smaller sample and use it to reduce standard error.
Proportionate stratified sampling
takes the same proportion (sample fraction) from each stratum
Disproportionate stratified sampling
takes a different proportion from different strata. This may be done to ensure minorities are adequately covered.
When is cluster sampling used?
Used when the studied population is spread across a wide area such that simple random sampling would be difficult to implement in accessing the selected sample
When is Snowball sampling used?
When you do not have access to sufficient people with the characteristics you are seeking
What is the parameter?
Characteristic of interest in the entire population, it is fixed value, e.g. Mean cholesterol level of all Asian males in Melbourne
What is the bernoulli distribution used for?
Only sample sizes of 1 so rarely used.
What is a binomial distribution?
Two mutually exclusive events. Probability of interest unknown (pi). The number of times an outcome repeats itself in data set.e.g. how many times did you role a 6 on a dice?
What is another name for the ‘normal distibution’?
Gaussian distribution
What type of distribution occurs when the mean < median
Negatively skewed
What type of distribution occurs when the mean>median>mode?
Positively skewed
What type of summary statistics do you use for normal distribution?
Mean and SD
What type of summary statistics do you use for asymmetric (left or right skewed) data?
Median and IQR
In the normal distribution where do 68% of values lie?
Within +- 1 SD (mean + or - 1SD)
In the normal distribution where do 95% of values lie?
+- 2SD
In the normal distribution, where do 99.7 % of values lie?
+- 3SDs
What is the z score formula?
Zscore= Observed value- True mean/True SD
What is the T distribution?
Very similar to normal distribution (symmetric and bell shaped) BUT bell peak of curve is FLATTER than normal distribution and tails are LONGER.
- Has df which are degrees of freedom
What is standard error?
The dispersion (spread) of the sampling distribution
- Measures how precise the population mean is estimated by the sample mean
(technically is the standard deviation of the sample mean)
When do we use the z score?
When the population (true) standard deviation is KNOWN
When do we use the T score?
When the Population (true) standard deviation is UNKNOWN and we only have the SAMPLE ST DEV
When would we use the standard deviation?
If we are describing the spread of observations in a study
When do we use standard error (SE)?
If using a sample to infer population results
e.g. Mean testosterone levels in elderly women in the population
As the sample size increases, what decreses?
Standard Error (SE)
If the SD is large then ____ is also large
If the SD is large, then SE is also large
What is the 68, 95, 99.7 rule?
- 68% of sample means expected to lie between 1SE of mean
- 95% of sample means are expected to lie between 2SE of mean
- 99.7 “”
What are the break downs for the distribution percentages (to add up to 99.7)?
- 34, 34, 13.5, 13.5, 2.35, 2.35 =99.7 (reference ranges)
What do we use if reference ranges (34,34 etc.) don’t apply?
- The Z or T score
What is a type I error?
When we reject the null hypothesis when it is true (false positive) e.g. telling man he is pregnant
Reject the null hypothesis when in reality the null is true – Type I error
What is a type II error?
When we retain the null when it is false (false negative) e.g. telling pregnant women she is not pregnant
What is the null hypothesis?
- Population mean 1- Population mean 2=0
What does variance equal?
- SD^2
What are the 3 assumptions for an independent T test?
- Two groups are independent
- People within each group are independent
- Birth weight in each group follows a normal distribution
What can the P value 0.05 help with in terms of type I and II errors?
- Fixing the type I error and minimising type II error or maximising power
What does a P value of 0.001 mean?
- this means that if 1000 similar studies were undertaken on the same population, only 1 out of 1000 studies would result in a sample result as extreme as the one obtained in this study due to sampling variability or by chance
- The study result is so rare that a chance factor can be ignored for the difference from the hypothesised value
In practice, we prefer higher confidence levels and narrower widths. A narrower width indicates smaller sampling variability and high precision. T/F
T
The width of a confidence interval can be reduced without reduction of confidence level by decreasing the sample size. (T/F)?
F
If the sample size increases, the standard error decreases which results in a narrower confidence interval.