Scribd
Upload
0 views

2

The document outlines the stages of preparation for collecting engineering data, emphasizing problem definition, study definition, and physical preparation. It details the importance of understanding the context, identifying response variables, managing influencing factors, and developing a data collection protocol. Additionally, it highlights the significance of assigning roles, conducting dry runs, and documenting pre-study predictions to enhance the data collection process.

Uploaded by

claire johnson
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
0 views

2

The document outlines the stages of preparation for collecting engineering data, emphasizing problem definition, study definition, and physical preparation. It details the importance of understanding the context, identifying response variables, managing influencing factors, and developing a data collection protocol. Additionally, it highlights the significance of assigning roles, conducting dry runs, and documenting pre-study predictions to enhance the data collection process.

Uploaded by

claire johnson
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 15

Preparing to Collect

Engineering Data
Math 89: Engineering Data Anslysis
Stages of Preparation
PROBLEM DEFINITION
1. Identify the problem to be addressed in general terms
2. Understand the context of the problem.
3. State in precise terms of the objective and scope of the study. (State
questions to be answered)
Stages of Preparation
STUDY DEFINITION
4. Identify the response variable(s) and appropriate instrumentation.
5. Identify possible factors influencing responses.
6. Decide whether (and if so how) to manage factors that are likely to
have effects on the response(s).
7. Develop a detailed data collection protocol and timetable for the
first phase of the study.'
Stages of Preparation
PHYSICAL PREPARATION
8. Assign responsibility for careful supervision.
9. Identify technicians and provide necessary instruction in the study
objectives and methods to be used.
10. Prepare data collection forms and/or equipment. Do a dry run
analysis on fictitious data.
11. Do a dry run analysis on fictitious data.
12. Write up a “best guess” prediction of the results of the actual study.
Step 1: Identify the Problem in
General Terms
Engineers prioritize problems based
on limited resources and job
responsibilities
Problem selection depends on
perceived importance and
solvability, often requiring team
consensus.
Data analysis, like Pareto diagrams,
helps quantify problem impact and Example of a Pareto Diagram

reduce subjectivity.
Step 2: Understand the
context of the problem.
Understanding the Context is Key: Engineers
must first thoroughly understand the process, its
inputs, environment, and relevant theory before
deciding what data will be helpful.
Flowcharts Aid Understanding: Creating
flowcharts of both the ideal and current
processes helps engineers visualize the system
and identify areas for improvement.
Organized Data Identification: Developing
flowcharts also provides a structured way to
identify data needs and potential variables of
interest for the statistical study.
Steps 4 and 5: Identifying the
response variables and possible
factors
• Define Response Variable: Precisely define the property of interest
and its measurement (e.g., "throughput" as pieces/hour, profit/hour,
etc.).
• Address Measurement Issues Early: Consider measurement
limitations and equipment needs before planning the entire study to
avoid budget or resource issues.
• Identify Influencing Factors: Leverage expert knowledge (including
hands-on experience and brainstorming) and tools like cause-and-
effect diagrams to systematically identify variables that may affect the
response.
Cause and Effect Diagram - What is a Cause and Effect Diagram and
How to Make One (Smartdraw)
Fishbone Diagram Examples, Importance and Purpose(Intellspot)
Step 6: Decide whether (and if so how) to manage
factors that are likely to have effects on the
response(s).
• Variable Selection: Using a list of potential influencing variables, their
suspected importance, measurement knowledge, and resource/time
constraints, decide which variables to actively manage in the study.
• Experiments vs. Observation: Evaluate the benefits of experiments
(control over variables) against the added costs and complexities
compared to observational studies.
• Manage Key Variables: Select a feasible set of managed variables so
that the impact of unmanaged variables doesn't overshadow the
effects of the variables of primary interest.
Step 7: Develop a detailed data collection protocol
and timetable for the first phase of the study.

• Setting Experimental Levels: Choose representative levels for


controlled/block variables (e.g., hydraulic fluid contamination)
reflecting typical operating conditions
• Detailed Protocol: Develop a comprehensive protocol including
experimental order (randomized to reduce bias), clear instructions for
data collection (so anyone can follow them), a timeline, and the
understanding that iterative data collection and analysis may be
required to reach a solution.
Experimental Design in Statistics (Calcworkshop)
Steps 8-9: Assignment of Roles and
Responsibilities
• Assign Data Collection Oversight: Designate a specific team member to
directly supervise data collection to ensure consistency and adherence
to the protocol. This helps prevent deviations from the plan.
• Technician Training: Provide thorough training to technicians involved in
data collection. This training should cover the study objectives and the
specific data collection methods to ensure accurate and reliable data.
• Emphasize "Why": Explain the reason behind the data collection to the
technicians. When they understand the purpose, they are more likely to
collect higher-quality information and follow the procedures correctly.
Steps 10-11: Dry Run (Collection
and Analysis)
• Data Collection Forms & Dry Run: Carefully designing data collection
forms and conducting a "dry run" analysis with fake data allows for
the identification and correction of potential problems before actual
data collection begins, saving time and resources.
• Team Review & Refinement: The dry run provides an opportunity for
the entire team to review the process and ask critical questions like,
"Will this data actually answer our research question?" This
collaborative review often leads to valuable refinements and can even
save a flawed study.
Step 12: “Best Guess”
• Pre-Study "Best Guess": Before conducting the study, document a
"best guess" about the anticipated results. This provides a baseline for
comparison.
• Demonstrate Learning: Comparing the pre-study guess to the actual
findings after the study is complete clearly highlights what was
learned and the value of the study itself.

You might also like