QMB project (2 Viewers)

[hongco1990]

Hi guys,
Im currently struggling with the QMB report and have no clue of what to do with the data set given. Are we supposed to start the project from week 2 or 3? Cos I just opened the file and dont know where to start with. Unlike QMA, theres no questions and the instruction sheet is like shit. Im just feeling lost in QMB
Hope you will give me a hand with it. Thanks a million.

 

[chewy123]

lol i havn't even started...
I think they have given the questions in italic, something about whether the hotel achieved their target...i will look into it later.

 

[TehAzner]

lol i havn't even started...
I think they have given the questions in italic, something about whether the hotel achieved their target...i will look into it later.

Might ask around 2nd year friends for you if you want. Everyone hates the QMA+QMB assignments lol

Btw Chewy enjoyed your mid-sesh break?? Reading week

 

[chewy123]

nah dw, i am doubtful our assignment is least similar><""

but yeh, hail reading week !! my break was good but too short lol, yourself?

 

[mr.nice.guy]

Is law srsly that bad???? thank god my uai was a few points lower

 

[hongco1990]

Im doing my project now (yea finally). Just a bit confused here with the regression. I wanna test the relationship btween "income" and other variables, however, income is qualitative variable (only have value of 0 or 1), so is it possible for a qualitative variable to have regression test? Cos in the textbook I read, the examples are all of quantitative variables

 

[Casmira]

yeah its called linear probability model. careful cause your SE are completely off when doing this kinda thing they need to be adjusted, not sure if thats in the scope of QMB though

if u wanna sound smart mention how it'd be better to use logit or probit model

 

[shakky15]

surely someone who has done this asignment could post up a copy?? so many people complete this subject each semester im sure theres one that could help out. im also stuck on where to begin.. and yeah most people say 'ask ur tutor' but ours can hardly fkn speak english lol... and i dont have the time to chase up head lecturers and what not since i work full time...

all ive really managed to do is run a statistic summary on excel..

 

[Casmira]

lol i dont blame u guys. qmb was total bullshit, the much harder stats courses had more helpful lecturers and they where easier than qmb so go figure good luck guys

 

[wrong_turn]

LPM can also veer off from the probability interval of 0 and 1.

 

[inneedofhelp]

Hi, can anyone post the instructions to the qmb project? I can seem to get my hands on them, webct not working

 

[whoisurdaddy]

how do u find the population means and standard deviation using the sample information? We cant use confidence intervals since we are not sure of the population sd. Can someone help?

 

[hongco1990]

how do u find the population means and standard deviation using the sample information? We cant use confidence intervals since we are not sure of the population sd. Can someone help?

Use the descriptive summary feature in Excel to produce sample mean and sample variance. Because the population variance is unknown, were not sure if population is normally distributed, so can use t-distribution. But as n>120 (ie 200 in this case), t-distribution can be approximated as normal, so we can z score to calculate.
You can definitely use CI as CI just a way to show how much youre confident with your sample data in inferring the population

 

[hongco1990]

Hi,
when im running the regression for Income and expenditure, follow the equation
[FONT=&quot]Expenditure = β0 + β1 Income + [/FONT][FONT=&quot]ε[/FONT]
with income only takes value of 0 and 1
b0 = 207
b1 = 43.80
p-value= 5.71 x 10^-9

1. how can I interpret [FONT=&quot]b0 & b1[/FONT]?
2. What does p-value mean in this case?

[FONT=&quot][/FONT][FONT=&quot][/FONT]

 

[wrong_turn]

1.
b0 : holding all factors constant, expenditure is 207 (this means that when expenditure alone is considered by itself, exp=207.

b1: the effect of income on expenditure when all other facts are held constant, the change is 43.8 for an individual whose income is considered as part of expenditure. (in econometrics, your income coefficient is considered a dummy variable. i.e. 0 when not activated and 1 when it is activated). therefore, the expected value of expenditure will be exp=207 + 43.80

2. the p-value is almost zero :S...

man i dont remember the interpretation but you should really read the textbook for that :S dont be lazy

 

[chewy123]

1.
b0 : holding all factors constant, expenditure is 207 (this means that when expenditure alone is considered by itself, exp=207.

b1: the effect of income on expenditure when all other facts are held constant, the change is 43.8 for an individual whose income is considered as part of expenditure. (in econometrics, your income coefficient is considered a dummy variable. i.e. 0 when not activated and 1 when it is activated). therefore, the expected value of expenditure will be exp=207 + 43.80

Does this mean income being 1 instead of 0 yields a $43.8 increase in expenditure?

Thanks.

 

[wrong_turn]

yes there is a change in 43.8 when income is considered.

 

[hongco1990]

1.
b0 : holding all factors constant, expenditure is 207 (this means that when expenditure alone is considered by itself, exp=207.

b1: the effect of income on expenditure when all other facts are held constant, the change is 43.8 for an individual whose income is considered as part of expenditure. (in econometrics, your income coefficient is considered a dummy variable. i.e. 0 when not activated and 1 when it is activated). therefore, the expected value of expenditure will be exp=207 + 43.80

2. the p-value is almost zero :S...

man i dont remember the interpretation but you should really read the textbook for that :S dont be lazy

Well, I think you got me wrong here. It would be correct if you interpret it that way in case of multiple regression (exactly what it said in the textbook chap 18 - so Im not lazy at all lol). But mine is simple regression with dummy variable involve , so I finally figure out a way of interpreting it. See if mine is right or not...
b0: average exp of those income <60k
b1: difference bwteen mean exp of income <60k and >60k
p-value: not sure

 

[chewy123]

Hongco, small p-value indicates "an overwhelming evidence to infer that a linear relationship exist" - page 637

 

[whoisurdaddy]

Is KPI 1 binomial distribution?