Biometry Practice Questions
13.
Draw a "decision tree" for when to use a t-test, correlation, regression, anova, or chi-square test, based on the nature of the response and explanatory variables.
Sum of Squares: deviations of all observations from their mean, relates to the total variance of the observations Degrees of Freedom: sum of individual degrees of freedom for the sample, equals one less than their sample sizes Mean square error: determine whether factors are significant, obtained by dividiing the sum of squares of the residual error by the degrees of freedom. Represents variation within the samples.
In a 1-way ANOVA table, what is the relationship between Sums of Squares (SS), degrees of freedom (df), and Mean Square error (MS)? What is the relationship between SS within groups, SS between groups, and SS total?
1: Specify an appropriate null model 2: find the corresponding null distribution 3: Use the null distribution to calculate a p-value 4: Use the p-value to make a decisions regarding the plausibility of the null model. Must be calculable, with a distribution
What are the four basic steps of "null model hypothesis testing", and what are the two key attributes of a test statistic?
Standard deviation: the amount of variability or dispersion from the individual values to the mean Standard Error: measures how far the sample mean of the data is likely to be from the true population mean. Standard error is always smaller than the standard deviation
What is the difference between SD and SE?
Test Statistic: the result of a statistical test P-value: the probability of obtaining results at least as extreme as the observed results of a statistical hypothesis test, assuming that the null hypothesis is correct, the smaller the p-value means that there is stronger evidence in favor of the alternative hypothesis.
What is the difference between a test statistic and a p-value?
Confidence Interval: the probability that a population parameter will fall between two set values Prediction Interval: an estimate that a future observation will fall, with a certain probability, given what has already been observed Interquartile Range: the middle 50% of values when ordered from lowest to highest. find median or lower and upper half of data, these become Q1 and Q3. The IQR is the difference between Q3 and Q1.
What is the difference between confidence interval, prediction interval, and Interquartile Range?
Mean: the average, used when all numbers are added up and divided by the number of samples Median: the middle number of the data
What is the difference between mean and median (what are they used for, and when)?
Parametric tests: make assumptions about the parameters of the population distribution from which the sample is drawn, as example is that the population data are normally distributed Non-parametric: "distribution free" can be used for non-normal variables, no assumptions are made about the data.
What is the difference between parametric and non-parametric tests?
T-distribution: similar to normal distribution just has fatter tails, have higher kurtosis, probability of getting values very far form the mean is larger Normal distribution: taller curve, smaller tails
What is the difference between the t-distribution and the Normal distribution? What is similar?
Type I: rejection of a true null hypothesis, also can be called a false positive Type II: non-rejection of a false null hypothesis, also known as a false negative
What is the difference between type I vs II error?
The purpose of doing a multiple comparison is to control the overall significance level. Obtain a error rate in an analysis involving more than one comparison
What is the purpose of multiple comparisons corrections?
Transforming data prior to analysis is a way to make the data as "normal: as possible, this increases the validity associated with the statistical analyses
What is the purpose of transforming data prior to analysis?
11.
What should be used in graphs for illustrating variation (SD, SE 95%CI, etc...), and why?
