SAINT LOUIS UNIVERSITY

SCHOOL OF ENGINEERING AND ARCHITECTURE

DEPARTMENT OF MECHANICAL AND MECHATRONICS ENGINEERING

In Partial Fulfillment

of the Requirements for the Course

ME 3141: Methods of Research for ME

3646 12:30 – 1:30 MWF H205

INDIVIDUAL COMPREHENSIVE REPORT

COURSERA

By:

RELIQUIAS, ELLY JOHN W.

BSME III

Submitted To:

ENGR. MARK EMRIEL S. PARLAN

Course Facilitator

30 OCTOBER 2024
 Comprehensive Exploration of Statistical Methods in Mechanical Engineering

Week 1: Descriptive Statistics

A major branch of statistics, descriptive statistics, was introduced in the first week, which deals
with summarizing and organizing data. I discovered the difference between cases and variables:
the variables represent characteristics measured in data sets (like age and test scores), and cases
are units of analysis. The ability to create a data matrix—rows for cases and columns for
variables—simplified analysis and interpretation of variables.

We also delved into levels of measurement, categorized as:

• Nominal: Categorized without order.

• Ordinal: Meaningful order over data presented, without fixed intervals.

• Interval: Without a true zero, equal intervals.

• Ratio: Actual 'zero points' with equal intervals.

By understanding these levels, their nature was understood, and appropriate statistical techniques
were selected for the analysis of this data. After this, data presentation was discussed as one of
the most important things to communicate effectively. For illustrating trends and distributions,
we highlighted some tools, like the table and the graph, in particular, the bar chart and
histograms.

We investigated some measures of central tendency, such as the mean, mode, and median, which
represent a data set. This was supplemented with our investigation of measures of dispersion
(range, interquartile range (IQR), variance, and standard deviation), which are important in
understanding data variability. Z scores were introduced to allow the interpretation of data point
values based on the entire distribution.

At the same time, this module served to put forward the importance of descriptive statistics as a
base foundation for advanced analytics, improving my ability to analyze data and form my
analytical skills, understanding the attributes of data, and the related next actions for such
decisions. This is crucial knowledge for my research focused on how mechanical engineering
practices relate to industrial safety outcomes.
Week 2: Bivariate Analysis

The second week shifted to bivariate, and I learned about the relationship between two
variables. I focused on correlating coefficients, especially Pearson's r, to quantify the strength
and direction of a linear relationship. I understood that correlation does not mean causation; it's
all very well to know that you should interpret relationships carefully, but this sort of raising a
glass in celebration of ‘stating the obvious' gets tiresome.

To examine categorical variables, I employed contingency tables; for visualizing

relationships between two quantitative variables, I employed scatterplots. With this analytical
approach, we gained an intuitive knowledge of what once ran but was now hidden among the
data. The module presented Ordinary Least Squares (OLS) regression analysis illustrating how to
model dependent and independent variables. One of the critical things to understand was the r-
squared statistic, which was used to assess a regression fit as a high value for r-squared reflected
a strong model, whereas a low value was a warning of a possible mislead conclusion from this
regression.

This bivariate analysis is a firm basis on which I can benchmark which mechanical
engineering practices make the most impact in terms of frequency and severity of industrial
incidents. Using Pearson's r, I’ll be able to calibrate the correlations between safety mechanisms
and the accident rate, and I’ll run OLS regression to model the effect that different practices have
on safety protocols

Week 3: Probability Theory

In the third week, I tackled probability theory concepts and the uses of them in real life.
The foundation for all this came from the idea of randomness (this showed that random activity
pervades everyday life). I was therefore taught to identify the event, sample space, and random
trial together with other components of probability at large.

Tree diagrams were introduced to visually represent an occurrence of unexplained events,

resulting in probabilities for different scenarios. Next we did some set theory using set notation
to better explain complex probability relations. As an essential concept, conditional probabilities
emerged: knowing how one event impacts another event allows us to view event interactions.
Then I found Bayes’ Rule, a useful rule to reconfirm beliefs when new evidence is presented, so
as to get things as accurate as possible when we face uncertainty.

The theoretical underpinning for this week's lessons forms the basis of which I can apply
to the probability of industrial incidents at Philex Mining Corporation. Here I plan to use tree
diagrams and conditional probabilities to quantify the effectiveness of the safety measures by
mechanical engineers and use probabilistic analysis to manage risk.

Week 4: Probability Distributions

During our fourth week, we examined probability distributions and saw that they are
important in quantifying random phenomena. It highlighted that there are both discrete and
continuous random variables, and the values these random variables take must be distinguished.
Thus we explored cumulative probability distributions to gain some insights into how
probabilities may be understood as ranges and to visualize probability behavior well.

As the world develops new techniques for studying the normal distribution and its
properties, I find myself learning about it having gone through the normal distribution and its
properties, and their importance in statistical methodology for the reason that most statistics
techniques are based on things that look normal. In addition to that, the binomial distribution was
something I was already comfortable analyzing for discrete outcomes.

This module gives me the insights I use to model and evaluate the risks associated with
different mining industry engineering interventions. I apply normal and binomial distributions to
see how different failure scenarios are likely and address how mechanical engineering makes a
contribution to operational safety.

Week 5: Inferential Statistics and Sampling Methods

The fifth module covered inferential statistics, which connect sample data to a larger
population. Descriptive statistics give us general knowledge about the data, while inferential
statistics are needed to extrapolate findings to entire populations. I learned about some sampling
methods and realized how important it is to be able to obtain representative samples to make
accurate conclusions.
 Consistent with the week’s content, this week set up sampling distributions vs. the
population from which they are drawn and laid a foundation for the Central Limit Theorem by
discussing how larger sample sizes lead to approximately normal distributions even when the
source population is not normal.

With insights into sampling methods, I am able to sense how engineering practices shape
the way things get done in industry and how the impact of safety interventions is distributed
across the company. The significance of the impact of engineering decisions on safety
performance will be better understood in my research by understanding sampling distributions.

Week 6: Confidence Intervals and Population Estimation

Confidence intervals, which provide an estimate of population parameters from sample

data, were the subject of the sixth week of the lecture. With confidence intervals, researchers can
tell us what a range of plausible values for population parameter means or population parameter
proportions would be. A bit more explicitly, I learned how important it was to pick the right
widths of these intervals so that wider intervals were bad and narrower, more precise intervals
were better with larger samples.

The implications of this knowledge are directly applicable to my dissertation research, in

which I will apply confidence intervals to estimate the efficacy of safety protocols implemented
by mechanical engineers at mine sites. This allows me to go into the implications of the
confidence intervals and lends me greater precision on my conclusions.

Week 7: Hypothesis Testing

The course ended with statistical hypotheses, and significance testing considered the use
of hypotheses as a guide to research largely. The deciding process with inferential testing was
clarified by learning about the null hypothesis (H0) and alternative hypothesis (H1). I
appreciated the necessity of finding the right weight of Type I and Type II errors and the urgency
for strong hypotheses for accurate conclusions.

These concepts are applied to my research, allowing me to evaluate the efficacy of safety
measures that mechanical engineers use rigorously. With that, I can formulate and test
hypotheses about the engineering interventions that are most strongly related to the prevention of
disasters in the mining sector and substantiate these engineering practices.

Summary:

Finally, following my 7-week voyage through the course on statistical methods, I am

greatly more aware of how these methods can be applied to the field of mechanical engineering,
specifically in the context of my research on how to mitigate industrial disasters at Philex Mining
Corporation. It was one week that built on another with descriptive statistics to hypothesis
testing.

Descriptive statistics were foundational, so with them I began with skills to represent and
organize data effectively. And this was important to understand the kind of data I was going to
see in my research. Moving to bivariate analysis permitted an exploration of relationships
between variables and the quantification of the impact of engineering practices on safety
outcomes. I immersed myself in probability theory and distributions and how random this
randomness in industrial operations is and how to model the uncertainties. This knowledge is
essential to determining the risk associated with different engineering interventions. Sampling
methods and inferential statistics were investigated further to cement my ability to make valid
conclusions about larger populations from sample data and thus produce reliable and applicable
research findings. These courses offered me the tools to compute population parameters and
more rigorously judge the efficacy of safety measures put in place by mechanical engineers. The
realization of how the Type I and Type II errors can manifest has played a significant role in
posing a well-formulated hypothesis and coming to my conclusion with the appropriate
groundwork.

All in all, I found that these statistical methods are not just theoretical concepts but
working tools for my research and a deeper understanding of how one could improve safety and
avoid disasters in the mining industry. With this comprehensive study on statistical methods, I
was able to get equipped with the analytical skills it takes to make a logical judgment and
suggestion so that the interests can be focused to better serve the safety interests of the Philex
Mining Corporation.
References:
Coursera | Online Courses From Top Universities. Join for Free. (2024). Coursera.
https://www.coursera.org/learn/basic-statistics/home/week/1

Coursera | Online Courses From Top Universities. Join for Free. (2024). Coursera.
https://www.coursera.org/learn/basic-statistics/home/week/2

Coursera | Online Courses From Top Universities. Join for Free. (2024). Coursera.
https://www.coursera.org/learn/basic-statistics/home/week/3

Coursera | Online Courses From Top Universities. Join for Free. (2024). Coursera.
https://www.coursera.org/learn/basic-statistics/home/week/4

Coursera | Online Courses From Top Universities. Join for Free. (2024). Coursera.
https://www.coursera.org/learn/basic-statistics/home/week/5

Coursera | Online Courses From Top Universities. Join for Free. (2024). Coursera.
https://www.coursera.org/learn/basic-statistics/home/week/6

Coursera | Online Courses From Top Universities. Join for Free. (2024). Coursera.
https://www.coursera.org/learn/basic-statistics/home/week/7