QMB3250 Exam 3 UF
One-way ANOVA test null hypothesis - ANSM1=M2=M3=M4....
tests the difference between population means, so we are looking to see if the difference is not
equal
replicates formula - ANS# of experamental units
--------------------------
# of treatments
interaction null and alternate hypothesis - ANSHo: Interaction is not significant
Ha: Interaction is significant
advantages of studying two factors simultaneously in an experiment?... - ANSa) It is more
efficient to study two factors simultaneously rather than separately.
b) We can reduce the residual variation in a model by including a second factor thought to
influence the response.
c) We can investigate interaction between factors
multiple regression? - ANStwo quantitative variables predicting a third quantitative variable
two way ANOVA - ANStwo factors
one way ANOVA - ANSone factor
another name for residuals? - ANSerror
(i.e. DFE -> degrees of freedom error)
to find degrees of freedom total from an ANOVA table?... - ANStake the total degrees of
freedom and add 1
what does it mean if 0 is included in the confidence interval? - ANSthat there is a statistically
significant difference
What does N and n represent in Kruscall Walis test? - ANSN - total number in sample
n - number in each group
, what is a dead giveaway for multicollenierity? - ANSwhen you have an overall F statistic that
shows that you have a significant model, but the p-values state that none of the variables
contribute to the model
what does it mean if there is a squared predictor? - ANSthe shape of the line is quadratic
what do you do with the interaction variable when using the line equation?... - ANSy-hat =
intercept +/- X1(x1) +/- X2(x2) +/- interaction variable (x1)(x2)
Backwards Elimination - ANSFit Model > Enter Y and All Predictor Variables > Change box in
top right to Stepwise. Select
"Enter All", Change Stopping Rule to p-value threshold. Select Direction as Backward and then
click on Step.
forward selection - ANSFit Model > Enter Y and All Predictor Variables > Change box in top right
to Stepwise. Select
"Remove All", Change Stopping Rule to p-value threshold. Select Direction as Forward and then
click on Step.
stepwise all possible models - ANSFit Model > Enter Y and All Predictor Variables > Change
box in top right to Stepwise. Under
the red triangle select All possible models. Use the defaults on the next window and select
Okay.
YOU WANT THE LOWEST RMSE.
all possible models VIF value - ANSThis is a high value of VIF. Some of the variables needs to
be removed to avoid
issues with multicolinearity.___
Fit Model > Enter Y and All Predictor Variables. Click Run. On the output, near parameter
estimates, right click. Select columns > VIF.
THIS IS NOT A STEPWISE!!!!
Finding non-centered moving average by hand?.. - ANS(add the data numbers from all the
required periods together) / number of periods requested
for example -> (11+22+33+44) / 4
what is smoother, a simple moving average with period 5 or with period 10? - ANSperiod 10 -
when you smooth a longer range, more extremes get smoothed out
forcasting time periods using the non-centered moving average - ANSAnalyze > Specialized
Modeling > Time Series. Y is output and input is X. Red Triangle on the
One-way ANOVA test null hypothesis - ANSM1=M2=M3=M4....
tests the difference between population means, so we are looking to see if the difference is not
equal
replicates formula - ANS# of experamental units
--------------------------
# of treatments
interaction null and alternate hypothesis - ANSHo: Interaction is not significant
Ha: Interaction is significant
advantages of studying two factors simultaneously in an experiment?... - ANSa) It is more
efficient to study two factors simultaneously rather than separately.
b) We can reduce the residual variation in a model by including a second factor thought to
influence the response.
c) We can investigate interaction between factors
multiple regression? - ANStwo quantitative variables predicting a third quantitative variable
two way ANOVA - ANStwo factors
one way ANOVA - ANSone factor
another name for residuals? - ANSerror
(i.e. DFE -> degrees of freedom error)
to find degrees of freedom total from an ANOVA table?... - ANStake the total degrees of
freedom and add 1
what does it mean if 0 is included in the confidence interval? - ANSthat there is a statistically
significant difference
What does N and n represent in Kruscall Walis test? - ANSN - total number in sample
n - number in each group
, what is a dead giveaway for multicollenierity? - ANSwhen you have an overall F statistic that
shows that you have a significant model, but the p-values state that none of the variables
contribute to the model
what does it mean if there is a squared predictor? - ANSthe shape of the line is quadratic
what do you do with the interaction variable when using the line equation?... - ANSy-hat =
intercept +/- X1(x1) +/- X2(x2) +/- interaction variable (x1)(x2)
Backwards Elimination - ANSFit Model > Enter Y and All Predictor Variables > Change box in
top right to Stepwise. Select
"Enter All", Change Stopping Rule to p-value threshold. Select Direction as Backward and then
click on Step.
forward selection - ANSFit Model > Enter Y and All Predictor Variables > Change box in top right
to Stepwise. Select
"Remove All", Change Stopping Rule to p-value threshold. Select Direction as Forward and then
click on Step.
stepwise all possible models - ANSFit Model > Enter Y and All Predictor Variables > Change
box in top right to Stepwise. Under
the red triangle select All possible models. Use the defaults on the next window and select
Okay.
YOU WANT THE LOWEST RMSE.
all possible models VIF value - ANSThis is a high value of VIF. Some of the variables needs to
be removed to avoid
issues with multicolinearity.___
Fit Model > Enter Y and All Predictor Variables. Click Run. On the output, near parameter
estimates, right click. Select columns > VIF.
THIS IS NOT A STEPWISE!!!!
Finding non-centered moving average by hand?.. - ANS(add the data numbers from all the
required periods together) / number of periods requested
for example -> (11+22+33+44) / 4
what is smoother, a simple moving average with period 5 or with period 10? - ANSperiod 10 -
when you smooth a longer range, more extremes get smoothed out
forcasting time periods using the non-centered moving average - ANSAnalyze > Specialized
Modeling > Time Series. Y is output and input is X. Red Triangle on the
