PROCESS

QUALITY ENGINEERING

MEASUREMENT SYSTEM AND ANALYSIS

ASSIGNMENT – 1

on

BIAS

Submitted

by

SRINATH

THANGARAJRAJA

8151813

EXECUTIVE SUMMARY:

In this report, we are taking length

of an ignition tube to find whether the observation values are biased or not,

and the report is made by the values observed during the observation and the bias

value is found using the formulas in excel and graph from Minitab which are

shown to give clear results.

REPORT:

Example:

As a quality assurance engineer I was evaluating the

observations of the product and analysis of the measurement equipment indicated

that there should be no linearity concerns, so I had only the bias of the

measurement system evaluated. I took the measurement of the length of an ignition

tube in the motor bike which was measured by Vernier calliper in (mm)The

part was measured with reference to layout inspection to determine its

reference value. and was then measured 10 times to see the bias values. And the values of t statics and t critics are

found with standard deviation and standards error of the mean using the

observed value in excel.

Reference value =19.920

Trial

Observed

Value(mm)

Bias(mm)

1

20.7521

0.8321

2

20.8523

0.9323

3

20.7563

0.8363

4

19.999

0.079

5

20.9214

1.0014

6

19.8923

-0.0277

7

20.7991

0.8791

8

19.9976

0.0776

9

20.4464

0.5264

10

19.9812

0.0612

n

Average

Standard

Deviation,

Standard Error of Mean, b

Measured Value

10

20.4397

0.4254

0.1345

Reference

Value = 19.920 ?

=.05

t

statistic

df

Significant t value (2-tailed)

Average Bias

Confidence Interval of the Bias

Lower

Upper

Measured

Value

3.863

9

2.14479

0.519

0.2154

0.824

GRAPHS FROM MINITAB:

`

Conclusion & result:

From these graphs the value shows us that’s

the zero falls within the interval of bias so that I can assume that the

measurement bias is acceptable assuming no need of additional variation. And

more over all the values are lesser than t-critic value showing us the product

falls under the rejection category. SO, the

system is biased, and if the measurement has to be non-zero bias it

possibly should be recalibrated to achieve zero bias or the instrument should

have been worn out which should be maintained. The bias study is performed to

verify whether the instrument is measuring offset in a significant manner, the

tests also reflects on the calibration state of instruments and establish the suitability

of instruments to measure the process, the bias study should also be performed

when the instrument is initiated into a new process.

The calculations of the standard

deviation, standard error of the mean, average bias, lower limit and upper

limit are done in excel worksheets which are attached with this assignment.