Term
|
Definition
| With a census, the whole population is sampled, versus a portion of the pop'n |
|
|
Term
| Why is taking a census too difficult |
|
Definition
| time, expense, tedious, impractical, need proper participation and volunteering, and we sacrifice quality for quantity. |
|
|
Term
| What are the three stages to sampling? |
|
Definition
1. Determine who/what to sample
2. Determine how you're going to choose these subjects
3. Determine how many you'll need to be confident in your findings |
|
|
Term
| Your test subjects must be _____ ___________ of your population of interest |
|
Definition
|
|
Term
| Give an example of inclusion and exclusion criteria |
|
Definition
Inclusion: Selecting individuals who you want/need to study/observe
Exclusion: ruling out individuals who would confound the study results |
|
|
Term
| What are the different population types as we go from what we want to test, to what we actually study? |
|
Definition
Target Population
Source Population
Sampling Frame
Sampling Units
Study Population |
|
|
Term
| What the broad types of sampling strategies in Epidemiology? |
|
Definition
| Non-probability, Probability, and Others (either non- or probability) |
|
|
Term
| What are the types of Probability Sampling? |
|
Definition
Simple Random Sampling
Systemic Random Sampling
Stratified Random Sampling
Cluster Sampling
Multistage Sampling |
|
|
Term
| What are the types of Non-probability sampling? |
|
Definition
Convenience
Judgment
Purposive |
|
|
Term
| What is Convenience sampling? |
|
Definition
| Sampling units are chosen b/c they're easy to get |
|
|
Term
| What is judgment sampling? |
|
Definition
| the investigator chooses what he or she deems to be representative of the population |
|
|
Term
| What is Purposive sampling? |
|
Definition
| Sampling units are chosen on purpose because of their exposure or disease status |
|
|
Term
| T or F: Non-probability sampling is a bad method |
|
Definition
false! It is good, it just needs caution
- can produce biased results if the subjects you select are not representative of the target population |
|
|
Term
| What are the pros to doing non-probability sampling? |
|
Definition
- relatively easy
- can be cheaper if you choose subjects based on convenience
- appropriate for homogenous population |
|
|
Term
| In simple random sampling, it is required that the _____ _____ is known |
|
Definition
|
|
Term
| T or F: In SRS, all individuals have an equal chance of being included |
|
Definition
|
|
Term
| How does systematic random sampling work? |
|
Definition
A sampling interval (say, every 5 cows) is computed, and the starting point is selected on a formal random basis.
Then, simple or systematic random sampling is conducted within each strata |
|
|
Term
| T or F: In St.R.S., the percentage sampled in each strata must be the same in all groups |
|
Definition
|
|
Term
| How does stratified random sampling work? |
|
Definition
The sampling frame is broken into groups (strata) based on some factor that is likely to influence the level of the characteristic being measured.
Then, simple or systematic random sampling is conducted within each strata |
|
|
Term
| What is cluster sampling? |
|
Definition
| The sampling unit is a group of individuals with things in common (herd, household, geographic region, etc.) BUT - the unit of concern is still the individual |
|
|
Term
| T or F: Cluster sampling may be probability or non-probability |
|
Definition
|
|
Term
| T or F: In cluster sampling, not all the sampling units are tested |
|
Definition
|
|
Term
| When is multistage sampling convenient? |
|
Definition
- When there are too many individuals ina cluster to obtain measurements on
- when the individuals in a cluster are so alike that measuring just a few will provide sufficient info
- if sampling is done randomly, it can be more cost efficient than other probability sampling meathods |
|
|
Term
| What is multistage sampling? |
|
Definition
| Similar to cluster sampling, though sampling takes place at the cluster/group level as well as the individual level. (i.e. first sample random groups, then randomly sample individuals from each of those groups) |
|
|
Term
| Main difference between multistage and cluster sampling? |
|
Definition
Multistage: Selection occurs at two levels, and only a proportion of individuals are selected and measured
Cluster: All individuals in each group are selected and measured |
|
|
Term
| What is the sampling unit in cluster sampling? |
|
Definition
| The group of individuals (the herd, the barn, etc) |
|
|
Term
| What is the unit of concern in cluster sampling? |
|
Definition
| The individual (i.e. cow, child, etc.) |
|
|
Term
|
Definition
| It is how tight your CI is around your study mean/proportion |
|
|
Term
| On what does precision depend? |
|
Definition
| Sample size, variability of characteristics, and sampling strategy used. |
|
|
Term
| What does a smaller CI say about the precision |
|
Definition
| A smaller is range is more precise, and therefore better |
|
|
Term
| How do you calculate sample size for estimating simple characteristics? |
|
Definition
1. Estimate variance
2. Select level of confidence that your estimate will include the true value in the population
3. Specify desired precision
4. Use appropriate formula |
|
|
Term
|
Definition
The variance (of proportions)
where p = diseased inds
where q = non-diseased inds
(p = 1-q) |
|
|
Term
| what is the variance of the means? |
|
Definition
|
|
Term
| What is the Z-value for a 95% confidence interval? |
|
Definition
| 1.96 (for a two-sided hypothesis) |
|
|
Term
| What is another way to describe precision? |
|
Definition
|
|
Term
| In a descriptive study, why do we calculate sample size? |
|
Definition
| To estimate the mean or a proportion (different equations) |
|
|
Term
| If you decrease the allowable margin of error, what will happen to the size of the sample that you require? |
|
Definition
| It will drastically increase |
|
|
Term
| How do you calculate sample size in analytic observational studies? |
|
Definition
1. State null & either 1- or 2-sided alternative hypothesis
2. Determine what it is you're comparing
3. Determine how much of a difference b/w the groups you want to detect and if necessary, the expected variance
4.Set alpha and beta (confidence and power)
5. Use the appropriate formula or table (based on what you're comparing) to estimate sample size |
|
|
Term
|
Definition
| (P) The ability of a test to detect differences between groups when a real difference exists |
|
|
Term
| What is a commonly used power, and what is the Z-value? |
|
Definition
80% power, beta = 0.2
Z(0.2) = -0.84 |
|
|
Term
|
Definition
| When we reject the null, given that it's true |
|
|
Term
|
Definition
| When we don't reject the null, given that it's false |
|
|
Term
| What is the probability of making a type 1 error? |
|
Definition
|
|
Term
| What is the probability of making a type 2 error? |
|
Definition
|
|
Term
|
Definition
|
|
Term
| What is a twp sided hypothesis? |
|
Definition
| When you predict that a value can either be higher or lower in one group over another |
|
|
Term
| As the size of the difference between 2 means or proportions decreases, the required sample size.... |
|
Definition
|
|
Term
| As the level of power desired to detect a difference between the two groups increases, the required sample size.... |
|
Definition
|
|
Term
| As the number of confounders you're controlling for increases, the required sample size.... |
|
Definition
|
|
Term
| As the number of hypotheses tested increases, the required sample size.... |
|
Definition
|
|
Term
| What is a better method of representing the population that you want to characterize, prob sampling or non-prob sampling? |
|
Definition
| Prob sampling, though it may not always be practical or possible |
|
|
Term
| What is required to calculate an appropriate sample size? |
|
Definition
- desired precision of your estimate
- expected variation in the data
- (in analytical studies) the size of difference you want to detect b/w groups; and the desired power |
|
|
