Research Methodology - - Unit 7 - Week 3 - Data analysis and Modelling skills Page 1 of 5

[email protected]

NPTEL » Research Methodology

Completed
Course outline

About NPTEL

How does an NPTEL online course work?

Pre-requisite Assignment

Week 1: Group Discussion on Research

Week 1: Overview of Research

Week 2: Literature Survey, Experimental skills

Week 3 - Data analysis and Modelling skills

Data analysis Part 1

Data analysis Part 2
Modelling skills Part 1
Modelling skills Part 2
Text file of data analysis
Safety in laboratory
Week 3 Feedback Form: Research Methodology
Quiz: Assignment 3

Week 4: Technical presentation and writing

Week 5: Creativity in research

Week 6- Intellectual property

Week 7- Design of Experiments

Week 8- Department specific discussions

Text Transcripts

Books

mhtml:file://C:\Users\User\Desktop\Nptel\Research Methodology - - Unit 7 - Week 3 - ... 3/20/2025

Research Methodology - - Unit 7 - Week 3 - Data analysis and Modelling skills Page 2 of 5

Problem Solving Session - Jan 2024

mhtml:file://C:\Users\User\Desktop\Nptel\Research Methodology - - Unit 7 - Week 3 - ... 3/20/2025

Research Methodology - - Unit 7 - Week 3 - Data analysis and Modelling skills Page 3 of 5

What is usually reported after a parameter estimation exercise?

Value of the estimate.

Upper bound on the parameter.
Standard deviation of the estimate.
Difference between the estimate and true value.

Yes, the answer is correct.

Score: 1
Accepted Answers:
Value of the estimate.
Standard deviation of the estimate.

1 point
The signal-to-noise ratios for two different data sets D 1D1 and D 2D2 obtained for a process under
identical operating conditions are 25 and 100, respectively. Then, which of the following is / are true?

Errors in parameter estimates obtained from these data sets will be approximately in the ratio of 4:1.
Signal amplitude in D 2D2 is two times higher than in D 1D1.
Noise is of a larger power in D 1D1 than in D 2D2.
Estimates obtained from D 1D1 will have larger uncertainty than those from D 2D2.

Yes, the answer is correct.

Score: 1
Accepted Answers:
Noise is of a larger power in D 1D1 than in D 2D2.
Estimates obtained from D 1D1 will have larger uncertainty than those from D 2D2.

1 point
Which of the following is true concerning bias in parameter estimates?

It is the maximum possible difference between the estimate and true value.
An estimator is said to have zero bias if the average of estimates across all possible experimental
records is equal to the true value.
A non-zero bias implies systematic error in estimation.
The bias of a biased estimator can decrease with the sample size.

Yes, the answer is correct.

Score: 1
Accepted Answers:
An estimator is said to have zero bias if the average of estimates across all possible experimental records
is equal to the true value.
A non-zero bias implies systematic error in estimation.
The bias of a biased estimator can decrease with the sample size.

n2 s2
Measurements of two oscillatory signals x [k] = si sin2(ω0 ω0k) and y =co cos2(ω0 ω0k) are
1 point

obtained using two different sensors (S1 and S2) over a time interval. Then, identify the correct statement(s):

Sum of the readings of x [k] and y [k] is always necessarily unity.

The measurements from S1 and S2 are always non-negative valued.
When connected to an oscillation detection (e.g., spectrometer) instrument, it can show x [k] to have
frequency 2ω0 ω0.

Measurements of x [k] can be used to estimate readings of y [k] if the sensor S2 fails.

Yes, the answer is correct.

Score: 1
Accepted Answers:

mhtml:file://C:\Users\User\Desktop\Nptel\Research Methodology - - Unit 7 - Week 3 - ... 3/20/2025

Research Methodology - - Unit 7 - Week 3 - Data analysis and Modelling skills Page 4 of 5

When connected to an oscillation detection (e.g., spectrometer) instrument, it can show x [k] to have
frequency 2ω0 ω0.

1 point
Which of the following is true in estimation?

Errors in parameter estimates are usually inversely proportional to the sample sizes.

The true value of a parameter can never be obtained from a finite number of measurements (data points).
Curve fitting is always best done by least squares methods.
Parameter estimates can rarely be zero-valued.

Yes, the answer is correct.

Score: 1
Accepted Answers:
Errors in parameter estimates are usually inversely proportional to the sample sizes.
The true value of a parameter can never be obtained from a finite number of measurements (data points).
Parameter estimates can rarely be zero-valued.

1 point
Identify the incorrect statement(s) among the following w.r.t. a non-linear function for y = f (x )

The function f (x ) should necessarily have a quadratic or higher powers of x .

First-order linear approximations of non-linear models using Taylor’s series-based expansion involve
only first-order derivatives of f (x ).
For some values of x , the non-linear function and its linear approximation can be identical.
Estimating f (x ) from data is only possible if at least three data points are available.

Yes, the answer is correct.

Score: 1
Accepted Answers:
The function f (x ) should necessarily have a quadratic or higher powers of x .
Estimating f (x ) from data is only possible if at least three data points are available.

1 point
Which of the following is true concerning empirical (data-driven) and first-principles models?

First-principles models can never be linear, whereas data-driven models can always be forced to be
linear.
Empirical models usually have better interpolation capabilites than extrapolation.
It takes at least hundred observations to obtain a reliable empirical model involving two parameters.
The parameters of a first-principles model have to be necessarily obtained through experiments.

Partially Correct.
Score: 0.67
Accepted Answers:
Empirical models usually have better interpolation capabilites than extrapolation.
It takes at least hundred observations to obtain a reliable empirical model involving two parameters.
The parameters of a first-principles model have to be necessarily obtained through experiments.
You have not attempted all the questions.
Click 'Cancel' to return to the questions, or click 'Submit' to proceed with the submission.

Submit Cancel
Atleast one question is required to be answered to proceed for submission.

OK
Please wait while the assessment is submitted.

mhtml:file://C:\Users\User\Desktop\Nptel\Research Methodology - - Unit 7 - Week 3 - ... 3/20/2025

Research Methodology - - Unit 7 - Week 3 - Data analysis and Modelling skills Page 5 of 5

mhtml:file://C:\Users\User\Desktop\Nptel\Research Methodology - - Unit 7 - Week 3 - ... 3/20/2025