Below is a list of things you should be able to do or know. The list is not complete, but is a good start. Please post about which ones you "get" and which you "don't get yet". With that info, people can post to each other and help them be successful, also help our review on Thursday. Remember, Monday is a BIG test, don't forget your machine.
*Using machine: make scatterplots, get LSRL, make residual plots
*Analyze residual plot and know what a residual even is
*Interpret slope, y-intercpet, r and r-squared in context of problem
*Use your lin-reg line to make predictions, and the limitations of the predictions
*Know how to analzye/look at computer output
*Are comfortable with re-expressing data
*Are able interpret "strange" points: outliers, influetials and leverage
Subscribe to:
Post Comments (Atom)
13 comments:
~ORIGINAL POST~
*Using machine: make scatterplots, get LSRL, make residual plots: GET
*Analyze residual plot and know what a residual even is: GET
*Interpret slope, y-intercpet, r and r-squared in context of problem: GET
*Use your lin-reg line to make predictions, and the limitations of the predictions: GET
*Know how to analzye/look at computer output: GET
*Are able interpret "strange" points: outliers, influetials and leverage: GET
*Are comfortable with re-expressing data: not quite get YET
~ORIGINAL POST~
... i deleted original post, i realized i should be more specific.
*Are comfortable with re-expressing data: not quite get YET
I am still having trouble grasping the concept of WHEN to use WHICH methods.
*Using machine: make scatterplots, get LSRL, make residual plots
*Analyze residual plot and know what a residual even is
*Interpret slope, y-intercpet, r and r-squared in context of problem
GET
*Use your lin-reg line to make predictions, and the limitations of the predictions
GET
*Know how to analzye/look at computer output
MOSTLY GET
*Are comfortable with re-expressing data
GET
*Are able interpret "strange" points: outliers, influetials and leverage
GET
*using machine:
-Make scatterplot:GET
-get LSRL: DO NOT GET!!!!!!!****
-make residual plot: GET
*analyze and understand residuals:GET
*Interpret slope, y-intercept, r in context of problem: GET
---r-squared: still DON'T know how to word it properly
*Use lin-reg to make predictions:GET
-know limitations: GET
*Analyze/look at computer output: Need to Review
*comfortable re-expressing data: YES
*Are able to interpret strange points: GET
****I don't understand what we do with LSRL or what it is. Please comment about it if you understand it...perhaps make new post. Thanks
I am not yet comfortable with LSRL and re-expressing data but I feel I understand the rest of the unit.
i dont know what LSRL is. everything i think i understand. practice problems of all of these things would be great.
I don't know what LSDL is, and I'm not too comfortable with the exact meaning of R2, but i think i'm getting there. everything else i have a pretty good understanding. but, i bet as soon as the test comes, my brain will go kapoosh. goodnight stats peoples! i shall be hearing from u soon! : )
I'm good with all of them except
*Interpret slope, y-intercpet, r and r-squared in context of problem.
So I would like to go over that one tommarow
I would say that I pretty much understand everything. The only thing that I am not too comfortable with is deciding which methods to use when re-expressing data. I could also benefit from some review on how to classify strange points.
By the way, unless I am mistaken, I am pretty sure that the LSRL is just the equation for the regression line. 8D
I need review on... All of the above.
No I'm just kidding BUT it would be useful to go over how to properly word
Interpret slope, Y-intercept, r and r squared in context of problem,
Know how to analyze/look at computer output
and re-expressing data.
Thats pretty much everything I'm not completely sure about.
I understand everything for the most part. The only thing that gets me sometimes is which technique to use for re-expressing data but after trial and error I find the correct one.
I understand everything except for I am a little uncertain on r squared. Why is r squared so useful??
r-squared is important because it tells how much of the variation in y is accounted for by the model...high r-squared is good cause it shows your model high high predictive powers.
Post a Comment