20 Statistics DataAnalysis ExpResultsINTaPRES

Types of Data Analysis: Techniques and Methods
There are several types of data analysis techniques that exist based
on business and technology. The major types of data analysis are:
Text Analysis
Statistical Analysis
Diagnostic Analysis
Predictive Analysis
Prescriptive Analysis
Data Analysis Process
Data Analysis Process is nothing but gathering information by using

proper application or tool which allows you to explore the data and
find a pattern in it. Based on that, you can take decisions, or you
can get ultimate conclusions.
Data Analysis consists of the following phases:
Data Requirement Gathering

Data Collection
Data Cleaning
Data Analysis
Data Interpretation
Data Visualization
one variable statistics x (x-x') ^ 2 y
12 9.55 27
9 0.01 21
8 0.83 19
12 9.55 27
6 8.46 15
7 3.64 17
11 4.37 25
6 8.46 15
10 1.19 23
5 15.28 13
12 9.55 27
70.91
variance 7.09
av 8.91 20.82
sample 11.00 11.00
variance 7.09 28.36
std 2.66 5.33
cooeficient of variation 0.299 0.256
conf cof 1.96 1.96
Standard Error 0.80 1.54
margion of error 1.57 3.01
upper 10.48 23.83
lower 7.34 17.80
median 9.00 21.00
mode 12.00 27.00
max 12.00 27.00
min 5.00 13.00
geometyric mean 8.52 20.17
range 7.00 14.00
quartile 1 6.50 16.00
quartile 3 11.50 26.00
quartile 4 12.00 27.00
skewness -0.11 -0.11
kurtosis -1.67 -1.67
Mean 8.91
Standard Error 0.80
Median 9.00
Mode 12.00
Standard Deviation 2.66
Sample Variance 7.09
Kurtosis -1.67
Skewness -0.11
Range 7.00
Minimum 5.00
Maximum 12.00
Sum 98.00
Count 11.00
Largest(1) 12.00
Smallest(1) 5.00
Confidence Level(95.0%) 1.79
cov(X,Y) = E(X-X') * E(Y-Y') ++, +-, -+, -- X Y
corr = cov(X,Y)/ sqrt(var(X) * var (Y) )
coorealtion cooeficinet is The normalized value of covariance

Standard error
confidence about parameters not statistics
confidence parameters is in the interval not a probability
Z, t, F values
p value area unde or above Z,t,F values
for proportions
rho = n'/n
sample size: solve sides of the confidence interval for n
for two samples
For the two confidence intervals to be nonoverlapping, the upper edge of the lower confidenc
be below the lower edge of the upper confidence interval:
er edge of the lower confidence interval should
CPU Time
3.1
4.2
2.8
5.1
2.8
4.4
5.6
3.9
3.9
2.7
4.1
3.6
3.1
4.5
3.8
2.9
3.4
3.3
2.8
4.5
4.9
5.3
1.9
3.7
3.2
4.1
5.1
3.2
3.9
4.8
5.9
4.2
CPU Time
Mean 3.90
Standard Error 0.17
Median 3.90
Mode 2.80
Standard Deviation 0.95
Sample Variance 0.90
Kurtosis -0.43
Skewness 0.20
Range 4.00
Minimum 1.90
Maximum 5.90
Sum 124.70
Count 32.00
Largest(1) 5.90
Smallest(1) 1.90
Confidence Level(95.0%) 0.34 3.56 4.24
large sample > 30
small sample < 30

12.4/13.1
COMPARING TWO ALTERNATIVES
Paired Observations like similar workloads on two different systems Unpaired Observations
test difference between then test as if one variable
A B
5.36 19.12
16.57 3.52
0.62 3.38
1.41 2.5
0.64 3.6
7.62 1.74
Observations samples for alternatives
1. Compute the sample means:
2. Compute the sample standard deviations:
3. Compute the mean difference:
4. Compute the standard deviation of the mean difference:
5. Compute the effective number of degrees of freedom:
6. Compute the confidence interval for the mean difference:
ANALYSIS, INTERPRETATION, AND VISUAL PRESENTATION OF EXPERIMENTAL DATA
THE LINEAR MODEL AND NULL HYPOTHESIS / hypothesis testing
SIGNIFICANCE TESTING
Descriptive Statistics & Inferential Statistics

Statistics
Linear Models
Sgnificance testing / confidence interval
Pictorial representations
Confidence intervals
Planned comparisons
Means of accounting for different sources of variance
Mathematical process models
Equivalence techniques - are standard in vision science,
Model Model Name
R = a + bX + gY Multiple Regression (Additive)
R = a + bX + gY + dXY Multiple Regression (Bilinear) y = b0 + b1x1 + b2x2 + ... + bkxk + e
R = a + bX + gY + dY2 Multiple Regression (Quadratic in Y)
R = a + bi + gj Two-Way ANOVA (Additive) y = q0 + qAxA + qBxB
R = a + bi + gj + dij Two-Way ANOVA with Interaction y = q0 + qAxA + qBxB + qABxAxB
y = q0 + qAxA + qBxB + qABxAxB + e
R = a + bX + gj One-Way ANACOVA (Additive)
R = a + bi + gj + dbigj Tukey’s one-degree-of-freedom interaction model
1x1 + b2x2 + ... + bkxk + e
AxA + qBxB
AxA + qBxB + qABxAxB
AxA + qBxB + qABxAxB + e

20 Statistics DataAnalysis ExpResultsINTaPRES

Uploaded by

Copyright:

Available Formats

20 Statistics DataAnalysis ExpResultsINTaPRES

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

20 Statistics DataAnalysis ExpResultsINTaPRES

Uploaded by

Copyright:

Available Formats

Types of Data Analysis: Techniques and Methods

Data Analysis Process

Data Analysis Process is nothing but gathering information by using

Data Analysis consists of the following phases:

Data Requirement Gathering

corr = cov(X,Y)/ sqrt(var(X) * var (Y) )

coorealtion cooeficinet is The normalized value of covariance

sample size: solve sides of the confidence interval for n

for two samples

large sample > 30

small sample < 30

THE LINEAR MODEL AND NULL HYPOTHESIS / hypothesis testing

Descriptive Statistics & Inferential Statistics

You might also like