Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd Logo
Upload

Welcome to Scribd!

  • 959 views

    Statistics Work

    Uploaded by

    BrittanyWilliams
    This document discusses key concepts in designing experiments: 1. It provides an example of an experiment to study unpopped popcorn kernels, identifying response variables, factors, and levels. Major sources of variation like brand and power could be difficult to control. 2. Replication allows estimating experimental error and factor means, while repeated measures tests each factor combination on a subject. The example compares testing soda brands and temperatures. 3. Randomization mitigates bias and ensures independent observations. 4. A sequential approach starting with screening experiments provides better understanding of factors and interactions than a single large experiment.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Statistics Work

    Uploaded by

    BrittanyWilliams
    959 views2 pages
    This document discusses key concepts in designing experiments: 1. It provides an example of an experiment to study unpopped popcorn kernels, identifying response variables, factors, and levels. Major sources of variation like brand and power could be difficult to control. 2. Replication allows estimating experimental error and factor means, while repeated measures tests each factor combination on a subject. The example compares testing soda brands and temperatures. 3. Randomization mitigates bias and ensures independent observations. 4. A sequential approach starting with screening experiments provides better understanding of factors and interactions than a single large experiment.

    Original Description:

    Statistics Work for Design of Experiment

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses key concepts in designing experiments: 1. It provides an example of an experiment to study unpopped popcorn kernels, identifying response variables, factors, and levels. Major sources of variation like brand and power could be difficult to control. 2. Replication allows estimating experimental error and factor means, while repeated measures tests each factor combination on a subject. The example compares testing soda brands and temperatures. 3. Randomization mitigates bias and ensures independent observations. 4. A sequential approach starting with screening experiments provides better understanding of factors and interactions than a single large experiment.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    959 views2 pages

    Statistics Work

    Uploaded by

    BrittanyWilliams
    This document discusses key concepts in designing experiments: 1. It provides an example of an experiment to study unpopped popcorn kernels, identifying response variables, factors, and levels. Major sources of variation like brand and power could be difficult to control. 2. Replication allows estimating experimental error and factor means, while repeated measures tests each factor combination on a subject. The example compares testing soda brands and temperatures. 3. Randomization mitigates bias and ensures independent observations. 4. A sequential approach starting with screening experiments provides better understanding of factors and interactions than a single large experiment.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 2

    Brittany Williams

    INSY7300 Homework 1
    29 August 2014
    1.1. Suppose that you want to design an experiment to study the
    proportion of unpopped kernels of popcorn. Complete steps 1-3 of
    the guidelines for designing experiments. Are there any major
    sources of variation that would be difficult to control?
    1. Recognition of and statement of the problem. In a given sample,
    how many kernels were
    2. Selection of the response variable. The response variables would
    the number of unpopped kernels and the total number of kernels in
    a bag.
    3. Choice of factors, levels, and ranges.
    a. Factors and levels:
    i. Cook time
    1. 2 minutes
    2. 3 minutes
    ii. Brand
    1. Act II
    2. Orville
    iii. Cook setting
    1. High power (80%)
    2. Low power (40%)
    1.8. What is replication? Why do we need replication in an
    experiment? Present an example that illustrates the difference
    between replication and repeated measures.
    Replication indicates an independent repeat run of each factor combination.
    Replication allows the experimenter to obtain an estimate of the
    experimental error and the true mean response for one of the experiments
    factor levels.
    You want to determine which soda combination is most refreshing: soda
    (Mountain Dew, Coke, Sprite) and temperature (with or without ice).
    Repeated measures experiment: Each respondent tries only coke with
    ice and provides a refreshment rating. Next they try coke without ice
    and provide a refreshment rating. This is repeated until each factor
    combination has been tested.
    Replication experiment: the respondent tries the 3 brands of soda with
    ice and without ice, in random order, and provides a refreshment rating
    for each combination. So the three sodas are treated in each
    temperature setting, so there are three replicates obtained.
    1.9. Why is randomization important in an experiment?
    Randomization is important to mitigate the introduction of systematic bias
    into the experimental results. Randomization also satisfies the requirement

    that observations or errors are to be independently distributed random


    variables.
    1.10. What are the potential risks of a single large, comprehensive
    experiment in contrast to a sequential approach?
    In a single large, comprehensive experiment, it is assumed that you know the
    objective and all of the important factors. However, if you do an initial
    screening experiment, analyze that data, and continue to experiment based
    off that information, you may have a better understanding of what the
    important factors are, the impact of interaction effects, etc.
    B. Describe how this course can help you in your work (or intended
    future career).
    I worked on a project to redesign an attribute of a medical device, so that the
    device would meet specific Japanese design requirements. The body of the
    device was to remain the same; only a single interior feature was to change. I
    came up with four design options, and tested a number of variables to
    determine the most optimized design. If I had already had this course, the
    DOE would have been better approached! I plan to re-enter industry so
    having a better understanding of DOE will definitely be beneficial for future
    projects.

    You might also like