SAS 4-3 Correlation Analysis

Sdílet
Vložit
  • čas přidán 22. 08. 2024
  • Describes how to make inference for a correlation value using PROC CORR.
    This video is part of a free, open, non-credit course available from the Department of Statistics at North Carolina State University. The course is meant to provide training to our current graduate students on the basics of SAS so that learning how to use SAS doesn’t hinder their learning in the courses that use SAS in our programs.
    Course Goals: This course is designed to demonstrate the basic use of SAS for reading, manipulating, and plotting data via SAS Studio. How to conduct a basic correlation and regression analysis is also discussed.
    By the end of this course students will be able to:
    - read and write SAS programs
    - import well-formatted raw data into SAS
    - complete common data manipulation tasks in SAS
    - produce numerical and graphical summaries in SAS
    - conduct a basic correlation and regression analysis
    If you'd like to join the course, enrollment information is available here:
    bit.ly/welcome...
    The full playlist of videos can be found here:
    bit.ly/SAS-pla...
    More about the NC State Statistics program can be found here:
    statistics.sci...
    More about our Online Master of Statistics and Online Graduate Certificate programs can be found here:
    statistics.sci...

Komentáře • 6

  • @cnne9871
    @cnne9871 Před rokem

    Hello, I keep trying to plot a scatterplot while carrying out my correlation, and I keep getting this error message "Scatterplot with more than 5000 points have been suppressed". Not sure how to navigate this.

    • @statisticallearninggroup1389
      @statisticallearninggroup1389  Před rokem

      That's a common issue with larger data sets! You can get around it by adding an option on the plots option (confusing right!). Something like this:
      PROC CORR DATA = mylib.data PLOTS(MAXPOINTS = NONE)=all;

  • @sabaaslam4766
    @sabaaslam4766 Před rokem

    Is it necessary to have the same number of observations for every variable to find the correlation?

    • @statisticallearninggroup1389
      @statisticallearninggroup1389  Před rokem

      Yes! Correlation is finding a measure of the linear relationship between two variables measured on the **same units**. If you don't have one of the two corresponding variables then you can't use those observations to measure the relationship.

  • @user-pp7qv4fz5z
    @user-pp7qv4fz5z Před 3 měsíci

    Where are the data sets?

    • @statisticallearninggroup1389
      @statisticallearninggroup1389  Před 3 měsíci +1

      You can join the course via the information in the details of the video. That will give you access. Alternatively, some of the same datasets are available at: www4.stat.ncsu.edu/~online/datasets/