Question

Spring 2019 Name, 33) If the 2 variables that we want to study are not mumeric, which method will we use to study their association? a) z2 table and calculations e) Normal distributions b) correlation and regression d) hon-linear correlation 34) If there is a high correlation between variables, the causality relationship between therm a) x causes y b) coincidence e) x and y are caused by a third variable(s) d) all of these are possibilities For items 35)-36): Suppose that a study of correlation between the number of cigarettes smoked per capita and death rate from lung cancer, we found thatr-0.7, and the regression line has an equation y-62 x + 38,where x-# of cigarettes s each day (average for the people of a given country) and y - death rate from lung cancer in deaths per million people in the population of the country. smoked death rates may be attributed to the linear b) 36% 35) what % of the variability in lung cancer relationship with smoking x cigarettes a day? d) 70% c) 64% a) 49% 36) What does the slope of the regression line above indicate? a) There is a high correlation between the avg, # of cigarettes smoked and death rate due to lung cancer. b) Regardless of correlation between variables, the line may be used to predict the lung cancer death rate c) That 62 people per million die from lung cancer in the country studied whether they smoke or not d) Each addi tional cigarette smoked, on avg, increases the death rate from lung cancer by 62 per milli 37) Predict the grade of a student(y) from the # of days absent frorn class (x) if the regression line equation is y -02x + 4.0 where the grades are as usual Dl.0 C-2.0 B-3.0 A-4.0, if the student missed 9 days ofclass. a) C+-2.8 b) С-22 c) B+-3.3 d) not passing 38) Would you trust the prediction above more if r -0.5 or-0.9? Why? a) trust r -0.5 because it is closer to 0 than-0.9 e) trust r- -0.9 because the points are closer to the line d) both predictions are equally trustworthy b) both are bad predictions becauser<0 39) What is used as a "point estimate" of the population proportion x? a) the sample mean b) the sample proportion p e) Confidence Interval d) none of these 40) Find the "expected value" for the number of days below 60 degrees in May in St. Paul a) The average number of cool days recorded for May in St. Paul over the past years b) We could calculate by (1 cool dayXprobability of I ed)+(2 c.d.) (prob. 2 c.d.)+ e) both a) and b) describe determining the "expected value" or long-term average d) none of the above 41) Calculate the standard deviation for the proportions of "B-students" if in a random sample of 88 MCTC students 50 got "Bs", and explain its meaning: a) we cannot calculate unless we take many samples 0.0528 is the std. dev. of the "B" students c) 0.57 is the standard deviation for this sample d) ơ ~ 0.0528 is the std.de0fps if we took many samples

Answer 1

33)

I would suggest to go for option (b) as for non-numeric data if it is ordinal then you can use the spearman's correlation and you can also use the option (a) (just because it is not mentioned that what type of data you have both (a) and (b) can be correct)but i would suggest you to go for (b) because using chi-square for association will tends to loose so much data and information you have collected.

option(d) is definitely not correct option because we use non-linear correlation when the association is not constant but it not like we use it for non-numeric data.

34)

option(d) is the correct option because all of the given points makes sense.

Causality means y occurs because of x. Ex- Egg is broken because it falls from desk.

35)

option (a) is the correct option.

r=0.7

r² = 0.49 (Coefficient of determinition)

So, there is 49% of variability is explained in our data we have.

36)

option (d) is the correct option.

Regression equation is y= 62x+38

Here, slope is 62 which indicates that each additional ciggerette smoked on an avg increases the death rate from lung cancer by 62 per million.