click below
click below
Normal Size Small Size show me how
_Statistics
Statisitcs
Question | Answer |
---|---|
What variable? Gender, Mortality, Race | Nominal: classified into groups in no particular order, no indication of severity |
What variable? Faculty evaluation, pain scales | Ordinal: Ranked in a specific order, but with no consistent level of magnitude difference between ranks |
What Variables? Classified into groups in no particular order, no indication of severity | Nominal |
What Variables? Ranked in a specific order, but with no consistent level of magnitude difference between ranks | Ordinal |
What variable? Fahreneheit | Interval |
What variable? Data are ranked in a specific order with a consistence change in magniture between units. Zero point is arbitrary | Interval |
What variable? Kevlin temp, Heart rate | Ratio |
What variable? Data are ranked in a specific order with a consistence change in magniture between units. Zero point is absolute | Ratio |
What variable? Controlled for in an analysis? | Independent variable (smoking) |
What variable? Measure as an outcome | Dependent variable (low birth weight) |
What is Type I Error? | Reject H0 and falsely accept H1. Find a differnce when there really isn't one. FALSE POSITIVE |
What is Type II Error? | Reject H1 and falsely accept H0. Fail to find a diffence when there really is one. FALSE NEGATIVE |
H0 (Null hypothesis) | State that there is no difference between group. H0 is negative statement (cow = human size) |
H1 (Research Hypothesis) | There is a difference |
Chi Squared | Norminal (Independent) |
Fisher's Exact | Norminal data (Independent) |
McNema | Norminal data (Dependent = Paired) |
Mann-Whitney U Test | Ordinal data (Independent = 2 group) |
Wilcoxon Rank Sum | Ordinal data (Independent = 2 group) |
Kruskal-Wallis | Ordinal data (Independent = 3 or more group) |
Wilcoxon Signed Rank | Ordinal data (Dependent = 2) |
Friedman | Ordinal data (Dependent = 3 or more) |
Student T | Internal/Ratio Data (Independent = 2) |
ANOVA | Internal/Ratio Data (Independent/Dependent = 3 or more) |
ANCOVA | Internal/Ratio Data (Independent = 3 or more) |
Paired Student | Internal/Ratio Data (dependent = 2) |
ARR | EER - CER |
RRR | ARR/CER |
NNT | 1/AAR |
Sensitiviy? | The ability of a test to reliably detect the presence of disease in an afflicted patient (TRUE POSSITIVE) |
Specificity? | The ability of a test to reliably detect the absence of disease in the unafflicted patient (TRUE NEGATIVE) |