Midterm

Question 1

Experiments vs observational studies

Answer 1

Exp. -> can put a control (témoin), researcher has control. Observational study -> we observe. We only associate and do not establish causality

Question 2

Counfounding variable

Answer 2

Variable that can affect both the treatment and the response

Question 3

Random sample

Answer 3

Representative of the population. Not biased

Question 4

3 types of bias

Answer 4

Selection bias. Non-response bias. Measurement errors

Question 5

Selection bias explanation

Answer 5

Subset of the experimental units of population is excluded/no chance of being selected for exp. = not a random sample

Question 6

Non-response bias explanation

Answer 6

Unability to obtain data on all experimental units selected for the sample

Question 7

Measurement error explanation

Answer 7

Inaccuracies in values recorded.

Question 8

Commonly used displays for qualitative data

Answer 8

Pie charts, bar plots

Question 9

Simpson’s paradox explanation

Answer 9

Third confounding variable changes the relationship between two other QUALITATIVE variables. Imbalance of the distribution of the categories of the third category with respect to the first two.

Question 10

Graphical displays for quantitative data

Answer 10

Boxplot, dotplot, histogram

Question 11

Adv/Disadv of the dotplot

Answer 11

A : Get to see all the data points + Easy to interpret D: Gets messy quickly if lots of data

Question 12

Adv/Disadv of histograms

Answer 12

A: Easy to pick up on all aspects of quantitative data (centre,spread, etc.) + made by most statistical packages D: Different bin widths can give diff. intepretations

Question 13

Mode

Answer 13

Number (or centre of the bin) that occurs most often (that has the most observations)

Question 14

Frequency vs Percentage

Answer 14

Frequency = number of observ in this bin. Percentage = percentage of total obs. it represents

Question 15

Different possibilities for the mode

Answer 15

Unimodal, Multimodal or no mode

Question 16

Adv/Disadv. of mean

Answer 16

A: Good for estimating pop. mean. + good inferential properties D: Outliers and skewed data

Question 17

Adv/disadv. of median

Answer 17

A: Easy to interpret + not infl. by outliers D: Bad inferential properties + longer to calculate

Question 18

Adv/disadv of mode

Answer 18

A: Highest concentration of data + bimodal data D : Bin width (class definition) matters

Question 19

sample vs pop mean symbols

Answer 19

X bar and μ

Question 20

Sample var/std vs Pop. var/std

Answer 20

S square or S vs sigma square or sigma

Question 21

Different measures of centre

Answer 21

Mean, median, mode

Question 22

Different measures of spread

Answer 22

Range, IQR, Var or std

Question 23

Something particular in variance

Answer 23

squared units

Question 24

% of observations with Z-score values 1,2.3 w/ empirical rule

Answer 24

68, 95, 99.7

Question 25

What are the three quartiles on boxplot

Answer 25

Q1 is 25th percentile, Q2 is median, Q3 is 75th percentile

Question 26

Where whiskers end on boxplot

Answer 26

At most extreme data point within 1.5 IQRs of the EDGE of the box in either direction

Question 27

Extreme outliers in boxplots

Answer 27

Anything beyond 3 IQRs of the EDGE of the box

Question 28

Ad/disadv. or variance

Answer 28

A : Good inferential properties D : Influenced by outliers and skewed data

Question 29

Ad/disadv. IQR

Answer 29

A : Not infl. by outliers D: Bad inferential properties and longer to calculate

Question 30

Ad/disadv. range

Answer 30

A : Easy to calculate D : Highly inf. by outliers and bad inferential properties

Question 31

Something particular we can see with boxplots

Answer 31

Possible to have no whisker on one side.

Question 32

Alternative formula for variance

Answer 32

1/n-1 times ((sum of squared Xi) - n*squared mean))

Question 33

Sensitivity (of a test ?) in prob

Answer 33

Pr (positive I disease)

Question 34

Specificity (of a test ?) in prob

Answer 34

Pr (negative | not disease)

Question 35

Probability of false positive

Answer 35

Pr (positive I not disease)

Question 36

Multiplicative rule

Answer 36

Pr union = condition * Pr (B)

Question 37

Limitation to using counting rules

Answer 37

Experience must haves an equal probability for each outcome