EmpowerStats

EmpowerStats
Project
Data
Analysis
Statistics
Tools

System initialization ......(can register or log in, but not analyzing data now)

EmpowerStats user registration

Email address:		Confirmation code:
Password:		Confirm password:

Email address:		Remember me
User password:

Confirmation code:
New password:
Confirm new password:

Open a project

Create new project

1. Create a new folder under Empower4Mac/Analysis/	Should be consist of A-Z and/or 0-9 only Do not use "databook", "webTools", "www" (reserved name) length less than 20
2. Place the data file into the folder	Should be consiste of A-Z and/or 0-9 only. Data type: .xlsx, .xls, .csv, .txt, .rdata, .sas7bdat, .xpt, .sav
Data won't upload to EmpowerStats server	All analysis takes place in users computer. Data and output files saved only in users computer. EmpowerStats takes no means to access your data and outputs files.

Data file:
Read rows:	If only read part of data, enter start row #	End row number
The folder name:

Select variables for read:

gaussian
binomial
poisson
Gamma
negbin

Breslow
Efron
Exact
Discrete

gaussian
binomial
ordinal
poisson

gaussian
binomial
multinomial

identity
logit
log
inverse
probit

Variable (and code) Labeling

Copy name(s)
Treat as continuous
Treat as categorical
Recode and regroup
Categorize
Z-score and percentile
Function transform
Change value
Hide
Delete
Show distribution

Variable manipulation

(If multiple,use VAR1+VAR2)

Cluster IDs

Finite population correction

Strata

Nest

Weight

Change value:

Wrong entry

Correct entry

Change outliers as NA (missing) (Check/edit the definition of outlier)：

	< Mean - * SD
	> Mean + * SD
	< % percentile
	> % percentile

Wrong entry (eg. >100, ==200, >"A", NA)：

Blank means NA

Save as new variable:

Recode and regroup:

Old code	New code		Use of Median

Save as new variable:

Categorize:

Method

New name

Additional request

Dichotomize

If, stratified

Tertiles

by:

Quartiles

Quantiles

Dichotomize by percentile

Ref.population for percentile(if):

% percentile

0:≤P90; 1:>P90

By cut points (comma separate)

eg. 3,8

0:≤3; 1:>3,≤8; 2:>8

Create segment variables

#：1,2,3

Calculate Z-Score and percentile for:

Method details

New name

Z-score

If, calculate Z-score stratified by:

	If, reference group for calculating Mean and SD:

If, enter Mean and SD
(If stratified, enter with same order and use comma as separator

Stratum:
Mean:
SD:

Percentile (Within 0-1)

If, percentile stratified by:

	If, reference group for calculating percentile:

Function transform:

Function

New name

log Log(select base)

^X AntiLog(select base)

1/X Inverse

√ Square root

X² Square

Rounding or approximation

Re-scale to to

Box-Cox transformation

Number to date (eg. 1999-1-12), origin date:

Number to string with sepcified e length: prefix:

Characters replace (...) = any contents within parenthesis; X...Y = any string start with X and end with Y.

replace with:

Search for text or number

Pre-text	Search for text or number	Post-text

	If found nothing, code as:

Date/time string: Format as

convert to number, origin date:

Extract: Year

Extract: Month

Extract: Day

Extract: Hour

Extract: Minute

Extract: Second

Extract: Day of week (Monday, Tuesday, Wednesday, ...)

Extract: fraction of year

Extract: fraction of day

Remove non-numeric character, clean to numeric

Unique code, (if) stratified by: Sort:

Sum within stratum, stratified by:

Counts of repeated value

Method

New name

Derived from a formula:

Calculate date difference: (/365.25 convert to AGE)

Variable name
Date format as:

Joint of two variables (At least one of them is categorical variable)

Conditional assignment

	Conditional expression (eg. SEX == 1 && AGE > 40)		Assignment(or expression)
if		then
else if
else if
else

Residuals and Predicted by regression models (.RESD,.PRED)

Independent(X):	s(X):smoothing, Comma separate (eg. SEX，s(AGE), EDU)
Outcome(Y):	Distribution:
Time:	(if use Cox), time for predict(option)：
Stratified by:	(if model for each stratum)
Subgroup for establish model:	(if use sub-group to establish model, eg.POP==1)

Calculate statistics from a list of variables

Variables name:

(Comma separate)

Statistics:

Mean	Median	N
SD	P75	Concatenate(eg.1_0_1)
Min	P90	All true(each 0/1 =1)
P5	P95	Any true(each 0/1 =1)
P10	Max	Mahalanobis distance
P25	Sum	Customer function

Output variables sufix:

.MEAN	.MEDIAN	.N
.SD	.Q3	.ADD
.MIN	.P90	.ALL
.P5	.P95	.ANY
.P10	.MAX	.MAH
.Q1	.SUM	.CMF

Group comparison (ANOVA or t-test)

Variable name:

(Comma separate)

Group number:

(Same order as variables, comma separate)

Statistics:

N、Mean、SD	ANOVABMS、WMS、P
Mean difference,95%CI

Output variables sufix:
.N.x, .MEAN.x, .SD.x
.BMS, .WMS, .P
.MD： .LOW.,.UPP.,.PADJ.

Regression of variables (Y1、Y2) with X (compare slope if has Y2)

Variables Y1:	(Comma separate)
Variables Y2:	(Same order as Y1, comma separate)
X:	(Variables or value of X, same order as Y1, comma separate)

Output variables sufix:
.BETA1, .SE1, .P1
.BETA2, .SE2, .P2
.DIFF: .BETA,.SE,.P

New var. code (Comma separate):

Enter the conditions

	Variable name		Value or variable (comma separate ='or')

Import/Export variable coding/labeling

Copy list
Paste append
Add a row
Delete
Delete all
Refresh

Repeat above
Repeat last
Increament
Clear all
Set smooth
Select this
Clear

	Description of selection step	Selection criteria:
1
2