Multiple Linear Regression Model (SPSS output) Part 1 of 4

[nellaivijay]

Nice one

[MsDamNi]

watch?v=GDze0TDUJ_o

[parancsmegtagado]

when the video came to salary, you said smg like "yes one would assume that if a player scores more has bigger salary" that means that we turned our correlation, doesnt it? i mean that means that salary is the dependent and scored points is the independent variable..?

[Foaman]

How do you give weight to independent vars? What would that be called?

[lamejorpersona1]

thank you for this. you explain it very well.

[liams18]

would anyone be able to tell me how I would be able to derive a sampling intensity from a linear regression, Im looking at the variation of the effect of silvicultural treatmenst on different trees and want to reduce my smalping size

[YFLOInternational]

My inclination is that these data schemes can also be tweaked and used to measure the maneuverability of various terrorist threat actors in 4-D battlespace... if again, we were to modify the parameters and data characteristics

[MilliVanilli2007]

I have a sample with 32 participants. is that enough to do a multiple regression?

[ccw0808]

Thanks for the video. I benefited a lot from your clear presentation. Really good for students from social science.

[sloggettful]

I am teaching Statistical and Computer Applications in Valuation of Real Estate at Seneca College. I just showed all four of your videos and the students were very pleased to see that I sought out other ways to get my point across. You not only do a great job but substantiate my teachings.

Thank you for the videos. One question "do you have any more videos planned" and I my lesson this week is concerning Comprehensive Model Building - Data Screening and Testing.

[ozkabron77]

What's the software you used??

thanks

[badassboyz2004]

@ozkabron77 SPSS?

[Mishkafofer]

I love the voice of the Professor, so soothing. Its like saying to me: Regression is so 1+1=2.

[SueMoseley]

Brilliant!

[rollsthepaul]

There have been many abuses of this technique. For example, investment equations sometimes use fudge factors that are linear correlation coefficients. The data is actually buck shot on the plot and correlates about the same as wind direction and the colour of my underwear. Some correlations are strong but so obvious that they are of no use(eg. points scored vs pay level). Quite often the database does not contain much of the key data such as biorhythms or personal highs and lows.

[statsprof]

Yes, this technique is greatly abused.

But, when looking at your ex, pay level, team owners are surely always comparing the range of pay levels vis-a-vis points scored in the league. A simple linear regression model on just that variable could be (is?) used by them.

The other vars do turn out to be sig (except for team points - c part II) and a larger analysis of the entire league & additional vars could allow owners (or hockey enthusiasts) to see if any of them are in fact related.

[amplifiedsz]

hello, ur tutorial on mlr is really helpful, can u plz tell me how this analysis on nhl would be helpfull for a marketing manager?

[statsprof]

thanks shiraz4me!

In the full model it turns out that pims DO have an inverse relationship...we see this in the next video...the coefficient for the pims variable is actually negative and significant...the outlier confounds the result displayed in the pims vs points earned scatterplot

the intuitive relationship that u point out does in fact turn out to be the case

kudos to the "I'm not a stat's guy" for pointing this out :) and thx for watching!

(ps only 3 Leafs players in the sample data)

[trevoreh]

Well done Professor,

As I said yesterday, I'm very happy to see a professor that uses all the technology and all types of learning styles to help students learn. Going into this course my only objective was to get the credit. But you made stats fun. Great video!

[statsprof]

Thanks trevoreh!  I appreciate the feedback. Stats *is* fun :)

[shiraz4me]

Well done! I'm not a stat's guy, but for penalty minutes wouldn't it be an inverse relationship?  I.e. the more pims, the less pts? Also thanks for explaining how poorly the Leafs are doing ... I suspect they're all the data points well below the line!