Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| element/set of elements that forms basis of selection. |
|
|
Term
|
Definition
| actual list of sampling unit |
|
|
Term
|
Definition
| element or aggregation of elements which the data is to be collected and/or analyzed. |
|
|
Term
| Simple random sampling design |
|
Definition
| most basic, every member of population has an equal chance of being chosen, but seldom used due to being labor intensive. |
|
|
Term
| Systematic random sampling design |
|
Definition
| (Interval random sampling)consists of a master list of members of the population, then setting a interval by dividing total # of units by present size of sample. Use is good for long lists, large samples, but assumes no order is present and the order is known as periodicity. Can be biased. |
|
|
Term
| Stratified random sampling |
|
Definition
| consists of classifying the elements, dividing the target population categoreis of homogeneous groups and then draws independent random samples from each strata/category. Uses a smaller sample size and takes into account the effect on dependent variables. |
|
|
Term
|
Definition
| similar to stratified but groups are naturally occurring. The clusters are chosen randomly and can have more than one stage. |
|
|
Term
|
Definition
| repetition of two sampling concepts (listing and sampling). |
|
|
Term
|
Definition
| incudes, t-test, p-test, chi-squared to conclude the sample was drawn is of significance and not by chance or coincidence. |
|
|
Term
|
Definition
| whether the findings are important and can be generalized and provide if the study was conducted again the results would be the same. |
|
|
Term
| Common uses of stat inference |
|
Definition
| test differences between measures of central tendency, of percentages, proportions, rates, association and duration and rate of change of trends. |
|
|
Term
|
Definition
| between mean rates and proportion |
|
|
Term
|
Definition
| analysis of variance, regression, slope and direction of trends. |
|
|
Term
|
Definition
|
|
Term
|
Definition
| zero - no differences not indicative of the population. |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| range of values around sample stat that the "true" or population value lies. lower range (narrow) and higher range (wide) |
|
|
Term
| Significance important to determine |
|
Definition
| to extent results can be attributed to the project, size/magnitude of differences, program costs and other support for quantitative results. |
|
|
Term
|
Definition
| objectivity, basis for judging, taking descriptive stats and putting into relevant terms |
|
|
Term
| Considerations for statistical testing |
|
Definition
| the risk, similarity of population, applicability of test to the data, and specific decisions about project. |
|
|
Term
| Statistical testing provides |
|
Definition
| relationships between activities/outcomes and generalization of results to similar times, places and situations. |
|
|
Term
|
Definition
| cost of collection, time, and hypothesized relationships and technical capability. |
|
|
Term
|
Definition
lower alpha value, lower risk of Type I
lower alpha value, more rigorous the test
increase alpha value (.01-.05) increase in type I error, decrease type II error |
|
|
