Week 10-14 Flashcards
Give an example: Nominal level of measurement
Ethnic group- Chinese
Give an example: Ordinal level of measurement
University grades- pass, credit
Give an example: Interval/Discrete level of measurement
IQ scores
Give an example: Ratio/Continuous level of measurement
Height
Give an example: Categorical Data
Nominal and Ordinal
Give an example: Continuous data
Interval and Ratio
Define: Continuous data
is an actual value of the measurement
Define: Categorical data
is the number of cases that fall into a category
What are the 3 measures of Central Tendency?
Mean
Median
Mode
Define: Median
is the score that divides a ordered set of scores into two equal halve
Define: Mean
the average
Sum of a set of scores, divided by the number of scores
What are the measures of Variability?
Range
Percentiles and Quartiles
Standard Deviation
Measures of Variability
Define: Variability
Dispersion or spread of scores
Define: Range
Difference between the highest and lowest score
highest score - lowest score = range
Measures of Variability
Define: Percentiles and Quartiles
Percentiles divide data into 100 equal portions
Quartiles divide data into 4 quarters
• Q1=25%,Q2=50%,Q3=75%, Q4=100%
Measures of variability
Define: Standard Deviation
It is the average difference between any score and the mean
- Important because it includes information on all scores
What is called the Point Estimate?
the mean or median
What is called the Measure of Variability?
the Standard Deviation or the Inter-quartile range
What is the purpose of Descriptive Results?
Concerned with organising and summarising information about a collection of actual observations
Descriptive Results
Name the 3 ways data is described?
1) By measures of central tendency2) By measures of dispersion (variabiltiy)
3) By measures of association
Descriptive Results
Name the methods of describing data
with:
1) Numbers- percentages, SD, central tendency
2) Tables/Figures- Bar chart, histogram, polygon, scattergram
Name the 2 Frequency Measures in epidemiology
Incidence
Prevalence
Define: Incidence
The frequency of new occurrences of disease, condition, or death in a defined population over a period of time
Define: Prevalence
The number of persons in a defined population who have a specified disease or condition at a point in time
(Understanding results- step 1)
Name the 2 types of Data results
Categorical and Continuous
(Understanding Results)
What are the 4 levels of measurement?
Nominal- categories only
Ordinal- categories and ranks
Interval or Discrete- categories, ranks and equal intervals between
Ratio or Continuous- all of the above, plus a true 0 point
(Understanding results- step 2)
What are the 2 types of Results?
Descriptive
Inferential
Define: Inferential Statistics (results)
are about what can be inferred from the sample to the population
Explain the purpose of Inferential Statistics
are necessary for answering questions for those beyond the sample
- cause, prevention, diagnosis, treatment, prognosis
How do we estimate Population characteristics from Sample data?
We use the normal curve as a model for making statistical assumptions and estimations
If results are Inferential, the results are about what 2 types of significance?
Statistical significance
Clinical significance
(Hypothesis Testing)
Define: Null Hypothesis (Ho)
proposes that there is no effect
(Hypothesis Testing)
Define: Alternative Hypothesis (Ha)
the opposite of the null hypothesis
Define: Hypothesis Testing
the process of deciding statistically whether the findings of an investigation reflect chance or “real” effects at a given level of probability
What are the 2 possible explanations for a positive outcome in a study?
- that the research hypothesis is correct
2. that the observed difference between groups occurred by chance
Define: Probability
is the likelihood an event will occur, given all possible outcomes
Define: p-value
probability due to chance
p=0.05
Explain: Statistical Significance (inferential results)
- hypothesis testing
- result is a probability value (p-value= 0.05)
- results are statistically significant if they are below 0.05
Explain: Clinical Significance (inferential results)
- provides an estimate we can use in real practice
- result is a confidence interval
- results must be above the MID to be clinically significant
Define: Point Estimate
a single number regarded as the most plausible value from the sample data
Explain: Estimation
will provide a result AND a degree of certainty
Define: Measure of Variability
turns the sample result (point estimate) into what we estimate for the whole population
What is the most important statistic to find in a results section?
measure of variability- what we estimate for the whole population
What is the common measure of variability used for estimation?
the Confidence Interval
Explain: Confidence interval
- the 95% suggests the degree of certainty of the estimation
Define: Minimal Important Difference (MID)
is the smallest worthwhile difference expected by a patient to proceed with the treatment
What is Clinical Significance?
is when the treatment effect (confidence interval) is equal or more than the MID
Define: Risk ratio
is a comparison of risks
What are the Difference Values in Continuous and Categorical data?
Continuous: difference value=0
Categorical: difference value=1
A result is clinically significant if…
- the point estimate is larger than the MID
- most of the confidence interval is larger than the MID
- a narrow confidence interval (narrow=precise)
(Tree/Forest Plots)
When is 0 used as the point of no effect?
when data isn’t a ratio
(Tree/Forest Plots)
When is 1 used as the point of no effect?
when data is a ratio
Explain with an example: Cohort Study
Prospective:
Pick people who smoke and people who don’t smoke
Follow participants over time- see who gets lung cancer
Explain with an example: Case-Control Study
Retrospective:
Pick cases of people who already have lung cancer and people who don’t
Go back in time and find out who smoked
When is a result statistically significant?
when the p-value of the results are below p=0.05
Define: Measures of Central Tendency
These are numerical indices that provide a quantitative summary of the centre of the distribution
(Mean, mode and median)
