Unit 10 Flashcards

(4 cards)

1
Q

Linear Regression Hypotheses

A

H0 : B = 0, (Explanatory) has NO effect on (response)
HA: B ≠ 0, (explanatory) HAS an effect on (response)

OR
>, (explanatory) has a POSITIVE effect on (response)
<, (explanatory) has a NEGATIVE effect on (response)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Conditions (Linear Regression)

A
  1. Independence
  2. Residuals are NORMAL
    –> Stat, calc4 THEN “resid” in normal quantile plot (#6)
  3. Random
  4. The two variables have a LINEAR relationship
    –> Use residual plot (resid) in Y of scatterplot, #1. No pattern = linear, yes pattern = nonlinear
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Calculations (Linear Regression)

A

ALWAYS T-test
–> DF = n - 2
When data is given, use calculator
–> enter in lists. Stat, tests, F: LinRegTTest OR G: LinRegTInt

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

T Calculations (Linear Regression)

A

T = (Slope)/Standard Error slope

Slope +/- critical value * standard error slope

How well did you know this?
1
Not at all
2
3
4
5
Perfectly