+ Reply to Thread
Results 1 to 2 of 2

Thread: Comparing measured and calculated solar data

  1. #1
    Points: 4, Level: 1
    Level completed: 7%, Points required for next Level: 46

    Thanked 0 Times in 0 Posts

    Comparing measured and calculated solar data

    I have created a program that outputs the average solar radiation available over a day for a specific location. I would like to statistically compare the data I have generated to a set of measured data so as to evaluate whether or not the model I have created is accurate in predicting solar levels.

    My statistical background is not that strong. At the moment I have tried doing an anova and repeated measures anova however, I am not sure if this is what I am after. Should I be trying to run a regression analysis instead?

    Thanks for any input

  2. #2
    TS Contributor
    Points: 18,889, Level: 87
    Level completed: 8%, Points required for next Level: 461
    CowboyBear's Avatar
    New Zealand
    Thanked 427 Times in 328 Posts

    Re: Comparing measured and calculated solar data

    ANOVA is for when you predictor variable is categorical. Here it seems like you have a continuous predictor variable (predicted solar radiation) and a continuous response variable (observed solar radiation), so a simple linear regression does seem more appropriate.

    However, depending on your goals you could even do something more simple. I.e., simply calculate the difference between the observed and predicted values, and then run some descriptive statistics. This would allow you to answer questions like:

    1) What is the average magnitude of error? In other words, is there a tendency to over or underestimate radiation?
    2) What is the average absolute magnitude of error? I.e., how far out do the estimates tend to be?

+ Reply to Thread


Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

Advertise on Talk Stats