SECTION A
1. In the 1920’s, Sir Cyril Burt investigated factors related to
“juvenile delinquency” in London. As part of his analysis
he compiled a table showing the incidence of juvenile
delinquency, and various measures of social conditions
(e.g. percentage of total population living in conditions
classified as overcrowding), for each of twenty-nine
London boroughs. Part of the data are presented in the
table below:
London Borough Juvenile Delinquency
(Per 10,000)
Percentage
Overcrowding
Finsbury 42 34
Holborn 36 20
Shoreditch 28 32
Bermondsey 23 23
St. Pancras 21 22
Southwark 18 24
Stepney 17 29
Battersea 16 12
Deptford 16 13
St. Marylebone 15 18
Westminster 15 10
Paddington 14 15
Bethnal Green 14 28
Islington 14 19
Hammersmith 13 14
Lambeth 12 13
Poplar 12 21
Kensington 12 17
Chelsea 12 14
Greenwich 11 14
Camberwell 10 13
Fulham 9 13
Woolwich 9 8
Hackney 8 12
Lewisham 7 5
City of London 5 7
Wandsworth 4 7
Hampstead 2 7
Stoke Newington 0 8
TURN OVER
(a) What is the relationship between the two variables (i.e.,
juvenile delinquency and overcrowding)? Specify the null
hypothesis (H0) and conduct a suitable test to determine the
significance of the relationship.
(b) On being presented with these data, a well-meaning police
Chief Inspector is quoted as saying: “The only way we will
ever reduce the amount of juvenile-related crimes in the worst
affected boroughs is by tackling the overcrowding problem.”
Is the Chief Inspector correct in making this conclusion?
Justify your reasons.
(c) Suppose that a decision is made that the percentage of
overcrowding should not exceed 10% in any newly developed
borough. What rate of delinquency might the urban planners
predict on the basis of the data presented above? You may
assume that any error associated with the measurement of
variable X (i.e. the “predictor” – percentage of overcrowding)
is negligible, and therefore it is permissible to conduct
a linear regression analysis on the data.
SECTION B
2. Design an experiment to determine whether variations in digit
span between speakers of different languages are attributable
to the word-length effect or to some other difference between
cultures.
3. It has been suggested that the signals of the short-wave cones
of the retina are delayed relative to those of the long- and
middle-wave cones. Describe how you would set about
testing whether this is the case.
TURN OVER
4. Delusions are a “first-rank” symptom of schizophrenia.
There is emerging evidence that many people in the typical
population report mild forms of delusional beliefs.
How would you design a questionnaire to test whether this
tendency reflects an underlying personality trait?