Scribd Logo
Upload

Welcome to Scribd!

  • 39 views

    Predict The Response Variable: Kazeem Adepoju, PH.D Wednesday, June 17, 2020

    Uploaded by

    Adesola Adepoju
    This document discusses predicting a response variable (quality rating) based on a predictor variable (clarity rating) for 366 university instructors. A linear regression model found that quality rating can be predicted from clarity rating. The document asks two questions: 1) What is the predicted quality rating for an instructor with a clarity rating of 3.5? 2) What is the expected mean quality rating for instructors with a clarity rating of 3.5? It explains that the point estimates for these two quantities are the same, but the intervals around them (confidence vs prediction intervals) differ based on whether the response is for a specific individual or the population mean.

    Copyright:

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

    Predict The Response Variable: Kazeem Adepoju, PH.D Wednesday, June 17, 2020

    Uploaded by

    Adesola Adepoju
    39 views15 pages
    This document discusses predicting a response variable (quality rating) based on a predictor variable (clarity rating) for 366 university instructors. A linear regression model found that quality rating can be predicted from clarity rating. The document asks two questions: 1) What is the predicted quality rating for an instructor with a clarity rating of 3.5? 2) What is the expected mean quality rating for instructors with a clarity rating of 3.5? It explains that the point estimates for these two quantities are the same, but the intervals around them (confidence vs prediction intervals) differ based on whether the response is for a specific individual or the population mean.

    Original Description:

    Undergraduate Regression course

    Original Title

    Lec3_Week2_Prediction

    Copyright

    © © All Rights Reserved

    Available Formats

    PPTX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document discusses predicting a response variable (quality rating) based on a predictor variable (clarity rating) for 366 university instructors. A linear regression model found that quality rating can be predicted from clarity rating. The document asks two questions: 1) What is the predicted quality rating for an instructor with a clarity rating of 3.5? 2) What is the expected mean quality rating for instructors with a clarity rating of 3.5? It explains that the point estimates for these two quantities are the same, but the intervals around them (confidence vs prediction intervals) differ based on whether the response is for a specific individual or the population mean.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    39 views15 pages

    Predict The Response Variable: Kazeem Adepoju, PH.D Wednesday, June 17, 2020

    Uploaded by

    Adesola Adepoju
    This document discusses predicting a response variable (quality rating) based on a predictor variable (clarity rating) for 366 university instructors. A linear regression model found that quality rating can be predicted from clarity rating. The document asks two questions: 1) What is the predicted quality rating for an instructor with a clarity rating of 3.5? 2) What is the expected mean quality rating for instructors with a clarity rating of 3.5? It explains that the point estimates for these two quantities are the same, but the intervals around them (confidence vs prediction intervals) differ based on whether the response is for a specific individual or the population mean.

    Copyright:

    © All Rights Reserved
    Download as PPTX, PDF, TXT or read online from Scribd
    Download as pptx, pdf, or txt
    of 15

    Predict the response variable

    Kazeem Adepoju,Ph.D
    Wednesday, June 17, 2020
    Topics:
    -Prediction and the associated intervals

    -Predict y vs. Predict E(Y|X=x)


    Ratings of 366 instructors at one large campus in the Midwest were collected.
    Students provided ratings for different aspects on 5 point scales (with 1 being the
    worst and 5 being the best). We will focus on the following two aspects: clarity &
    quality

    Research question:

    If I know an instructor’s clarity rating, what


    can I say about the instructor’s overall
    quality rating?

    ^
    𝑞𝑢𝑎𝑙𝑖𝑡𝑦
      =0.221+0.952 𝑐𝑙𝑎𝑟𝑖𝑡𝑦
    The sample data

    X = clarity, Y = quality

    3.5

    If a (specific) professor’s clarity rating is 3.5, what is his/her quality rating?

    What is the expected (mean) quality rating for professors with clarity rating
    equal to 3.5?
    X = clarity, Y = quality

    If a (specific) professor’s clarity rating is 3.5, what is his/her quality rating?

    What is the expected (mean) quality rating for professors with clarity rating
    equal to 3.5?

    is ______________. E(Y|X=3.5) is ______________.


    A. A known parameter A. A known parameter
    B. An unknown parameter B. An unknown parameter
    C. A random variable C. A random variable
    Recall:
    Point estimate:  

    The estimated quality rating for a (specific) professor who has clarity rating equal
    to 3.5 is 3.55.

    The estimated mean of quality rating for professors who have clarity rating equal
    to 3.5 is 3.55.
    (This is a random (This is a parameter, most
    variable) of the time unknown)
    formula
    estimator

    The point estimates for these two quantities are equal! What about the
    intervals?
    General format of an interval (two-sided)
    Distribution (Estimated)
    Point estimate multiplier /quantile standard error

    Old example: 95% confidence interval for u, the population mean

     0.975-quantile
    OR
    from
     
    Where S is the
    estimated standard
    A parameter will have a confidence interval
    deviation
    Confidence Interval for
    ● Find a point estimate:

    ● Identify the distribution of the difference

    If σ is unknown, use If σ is known, use

    t distribution multiplier N(0, 1) multiplier


    Prediction Interval for
    ● Find a point estimate:

    ● Identify the distribution of the difference

    If σ is unknown, use If σ is known, use

    t distribution multiplier N(0, 1) multiplier


    Assume
      is unknown, calculate a 95% interval when :

    Distribution (Estimated)
    Point estimate multiplier /quantile standard error

    CI for expected
      0.975-quantile from
    response value
    (a parameter)

    PI for actual   0.975-quantile from


    response value
    (a random variable)
    Exercise:
    Professor John’s clarity rating is 3.5, provide a 95% interval for his quality rating.

    Is this a confidence interval or a prediction interval?

      0.975-quantile from

    > qt(p = 0.975, df = 366 - 2)


    [1] 1.97
    > mean(Rateprof$clarity)
    [1] 3.5

     ´ =3.5
    𝑋
    𝜎^  =0.1828
    n=366
    Professor John’s clarity rating is 3.5, provide a 95% interval for his Quality rating.

      0.975-quantile from

    > qt(p = 0.975, df = 366 - 2)


    [1] 1.97

     ´ =3.5
    𝑋
    𝜎^  =0.1828
    n=366
    3.5

    You might also like