usmleasr
Forum Guru
Topics: 105
Posts: 970
|
can any one explain me about various statistical tests?( paired t test, chi square test etc) ....they stay as gray areas for me :?
|
| mjl1717
Forum Hero
Topics: 958
Posts: 5,465
|
FIRST: you have to know nominal means different groups,this, that or those.
ordinal means rank order--olympic medals ,class rank in med school
interval has to do withgradation or mean and standard deviation.
V A R I A B L E S
Second:
NAME OF TEST INTERVAL NOMINAL COMMENT
Pearson 2 0 is therea linear relation
Chi-square 0 2 Any # of groups
t-test 1 1 2 groups only
One way ANOVA 1 1 2 or more groups
Matched pair t test 1 1 2 groups,linked data
pairs,before &after
Analyze this for starters
___________________
Smell the coffee! "Is That an Osler move??"
|
| mjl1717
Forum Hero
Topics: 958
Posts: 5,465
|
This new format wont allow me to set up a table and I put in a subject EVERY time
___________________
Smell the coffee! "Is That an Osler move??"
|
| usmleasr
Forum Guru
Topics: 105
Posts: 970
|
sorry i am not able to follow :roll: :icon_shake:
|
| meg
Forum Guru
Topics: 62
Posts: 806
|
actually first aid 2004 (probably earlier editions too) and BRS behavioral mention about these tests and what they mention is what we need.
look for what they are asking to compare:-
if they are asking about means, then see whether you are comparing 2 or more than 2 variables/entities..
if 2 varialbe - then it is a t test
if 3 or more - then ANOVA
else if it is not about means, but say about percetages, or some other way of comparing, it is a Chi square
we dont need to go to indepth about what these individually mean.
Check out FA
|
| usmleasr
Forum Guru
Topics: 105
Posts: 970
|
thanks meg
|
|
| |
| | | | | | |
