Lab 1

7. [2 points, ½ for b, d, and e; ¼ for a and c] A public health researcher wants to know the
percentage of parents whose children are starting their first year of school who have had their
children given the flu vaccine. He randomly selects 5% of the students from every A.C.T. school’s
first-year intake, and asks the students’ parents whether they have had their child vaccinated
against the flu.

a) What is the sampling frame being used here? The list of first-year students in the A.C.T.
schools.
b) What kind of sample is being drawn? Stratified random sample, with the schools as
strata.
c) Does each child starting primary school have an equal chance of being selected? Why or
why not? Yes, because 5% of the student population in each school is being selected.
d) What is the population parameter that researcher is estimating? Percentage of
children starting first year of school who have been vaccinated against the flu.
e) To what population can he statistically generalize? Children starting first year of school in
the A.C.T. during the year of the study.

8. [2 points, ½ for each part] A study of attitudes and behaviour regarding driverless cars
investigated whether people would take over driving if the traffic around them exceeded the speed-
limit (the driverless cars are programmed to obey speed limits). All participants were licensed
drivers. Participants’ age and gender were recorded, and they were randomly assigned to one of two
conditions. In condition 1, they were simply asked whether they would take over driving if the
surrounding traffic exceeded the speed limit by 10 km/hr. In condition 2, they were placed in a
driving simulator with virtual surrounding traffic exceeding the speed limit by 10 km/hr, and had to
decide whether to take over the driving or not.

a) What is (are) the experimental variable(s)? Whether the participant is answering a

question or making a decision in the simulator.
b) What is the dependent variable? Whether or not the participant would take over the
driving.
c) Are there any nonexperimental independent variables? If so, what are they? Age and
gender.
d) Describe at least one potential extraneous attitudinal variable in this study. Valid
examples include the participant’s views about risk-taking, law-abiding, morality,
preference for being “in control”, etc.

9. Using SPSS to obtain descriptive statistics [2 points, 1 for each part]

In SPSS open the file called PSYC2009survey2015brief.sav. The variable named NFCSUM is a “need
for closure” scale, widely used to measure the extent to which people feel that they need certainty
and closure in their lives. Examples of NFC scale items are:

 I find that I am able to establish a consistent routine, enabling me to enjoy life.

 I am able to have a clear and structured mode of life.
 I generally succeed in avoiding unpredictable situations.
The score on each item is from 1 to 5, with a higher number indicating stronger agreement with the
item. There are 12 items, so the minimum possible score on NFC is 12 and the maximum is 60.

Use what you have learned about SPSS to answer these questions:

What are the mean and standard deviation of the NFC score for females and males? (Q2 is the
variable that recorded whether a person is female or male)

Female mean = 34.34 Male mean = 30.61

Female s.d. = 8.58 Male s.d. = 8.05

Briefly interpret what these statistics suggest about differences between males and females
regarding NFC.

The means suggest that females agree more than males do with the NFC items.

The standard deviations indicate that the genders are similar in their variability on this scale.

10. [2 points: ½ for each part]What level of measurement is each of the following?
1. The candidate a person votes for in a local election. Nominal
2. Degree of confidence in a person’s prediction as measured by a number they nominate on a scale from
0% to 100% (e.g., “I am 75% confident that my prediction is accurate”). Ratio
3. Number of Facebook friends a person has as a measure of how large their friendship network is.
Interval, because they may have friends who are not on Facebook
4. The scale used in this survey question:
"How confident do you feel about your knowledge of the impact the GST has on tertiary students?"

Not at all Slightly Moderately Fairly Very Completely

Ordinal

Lab 2
11. [2 points: ½ for each part] On which of these variables could we meaningfully use the mean,
median and/or mode?
1. 200 persons’ hand-squeeze strengths (in kilograms per square centimetre) mean, median and mode
2. 150 persons’ education levels (classified as less than year 12, year 12, undergraduate degree, or
postgraduate degree) median and mode
3. 100 ANU students’ degree programs mode
4. 100 persons’ ratings of whether they "strongly agree", "agree", "neither agree nor disagree", "disagree",
or "strongly disagree" with a political opinion statement. median and mode

1. [5 points, 1 point for each part] A recent survey of graduates with professional degrees reported
selected findings regarding whether graduates in Law, Medicine, and Engineering gained
employment in their profession within one year of graduation. In probabilistic terms, here is the
information:

P(employed within 1 year) = .74

P(Law) = .28
P(Medicine) = .40
P(Law and employed within 1 year) = .22
P(Engineer and employed within 1 year) = .20

Easiest way to do this is multiply the probabilities by 100 and set up a table with 100 people:

Degree
Medicin
Law e Engr.

Employed within 1 year 22 32 20 74

Not employed within 1 year 6 8 12 26

28 40 32 100

a) P(employed|Law) = 22/28 = .7857

b) P(not employed|Engr.) = 12/32 = .3750
c) P(Medicine|not employed) = 8/26 = .3077
d) P(Medicine or employed) = (40+22+20)/100 = .82
e) P(not employed and Engr.) = 12/100 = .12

2. [5 points, 1 point for each part] Here is an AIDS testing type of scenario again, but this time with
Hepatitis, in a population of 1,500,000 adults.

Positive Negative

Hepatitis 78400 1600 80000

No Hep. 71000 1349000 1420000

149400 1350600 1500000

a) What is the false positive probability? 71000/1420000 = .05

b) What is the false negative probability? 1600/80000 = .02
c) What is P(Hep|Positive test)? 78400/149400 = .5248
d) What is P(No Hep|Negative test)? 1349000/1350600 = .9988
e) What is P(Negative test|No Hep)? 1349000/1420000 = .95

Lab 3
 4. [3 points, 1 for computation, 2 for “why”] In Q.1 you were presented with an anti-
psychotic drug treatment that provides detectable improvement in 55% of clinically
psychotic people. If we give this treatment to 10 psychotic clients, what is the probability
that more than 8 of them will show improvement? Why does this probability differ from
the probability you got in Q.1?
The sum of the probabilities in Table A.1 from 9 to 10 for N = 10 and P = 0.55 is .02326.
This is smaller than the Q.1 probability because we have 10 people here instead of just 5.
Larger samples produce results that are closer to the mean, so extreme results like this
are less likely with larger samples.

5. [3 points] You’re in charge of a quit-smoking program that is known to have a 1/3 success-rate
(i.e., the probability that a person completing the program will quit smoking is 1/3). You have 30
people enrolled in this program, and so far 7 of the first 9 to complete the program have quit
smoking. What is the best estimate for the number of people out of the 30 who will have quit
smoking after completing the program?

We should expect the remaining 21 to have a probability of 1/3 of quitting, and 21(1/3) = 7. Given
that 7 already have quit, the expected number of quitters is 7+ 7 = 14.

6. [4 points, 1 each for the first two parts, 2 for the third] In the verbal test from Q.3., what score is
exceeded by the top 10% of the 9-year-olds, and what score is exceeded by the top 10% of the
10-year-olds? Explain the close similarity between the two scores, despite the difference
between the two age-groups’ means.

From Table A.2, the z-score corresponding to an area beyond z of .10 is z = 1.28. That’s the

number of standard deviation units below the mean.

So, the 10-year-olds’ score is 120 + (1.28)(15) = 139.2;

and the 9-year-olds’ score is 115 + (1.28)(20) = 140.6.

The scores are close because the 9-year-olds’ standard deviation is larger.