Login

Welcome, Guest. Please login or register.

April 19, 2024, 09:09:56 pm

Author Topic: Least Squares Regression Question  (Read 811 times)  Share 

0 Members and 1 Guest are viewing this topic.

BAH0003

  • Trailblazer
  • *
  • Posts: 33
  • Respect: 0
Least Squares Regression Question
« on: April 12, 2019, 09:06:40 am »
0
Hey guys,

I was working on this question and was having a bit of trouble getting to an answer. Is it okay if someone can explain this question and what the answer would be (it is B apparently)?

Thanks

RuiAce

  • ATAR Notes Lecturer
  • Honorary Moderator
  • Great Wonder of ATAR Notes
  • *******
  • Posts: 8814
  • "All models are wrong, but some are useful."
  • Respect: +2575
Re: Least Squares Regression Question
« Reply #1 on: April 12, 2019, 09:23:13 am »
+2
It's actually just simultaneous equations.
\[ \text{Because we're given some of the predicted values}\\ \text{on the actual regression line, we just sub them in.} \]
\[ \text{So if the regression line is }y=a+bx\text{, we have}\\ \begin{align*} 10.7 &= a+900b\tag{1}\\ 6.7 &=a+1700b\tag{2} \end{align*} \]
\[\text{If you have calculators that solve simultaneous equations then just plug it in}\\ \text{but otherwise, }(2)-(1)\text{ gives}\\ 800b = -4 \implies \boxed{b = -0.005}\]
(Note of course that the slope is \(b\).)

MEH0010

  • Adventurer
  • *
  • Posts: 18
  • Respect: +4
Re: Least Squares Regression Question
« Reply #2 on: April 12, 2019, 07:30:30 pm »
+1
Hey guys,

I was working on this question and was having a bit of trouble getting to an answer. Is it okay if someone can explain this question and what the answer would be (it is B apparently)?

Thanks
Hope this helps :)
*Answer is attached
« Last Edit: April 12, 2019, 07:32:36 pm by MEH0010 »