New A-level Biology

The Bay Sixth Form

Statistics booklet
A-level Biology Statistics booklet

The Bay House Sixth Form

A-level Biology Statistics booklet

1
 CONTENTS
About Statistics ………………………………….………….3
Descriptive Statistics ………………………………………3
Range…………………………………………………..……..………………..4

Standard deviation…………………………………………….………………5

±2Standard deviations………………………………………………………..6

Worked examples……………………………………………….6

Questions………………………………………………………...9

Standard Error of the mean and 95% Confidence Intervals….………....11

Worked examples…..……………………………..……..…….12

Questions………………….....………………..…….…………16

Inferential Statistics …………………………….………..21

Chi-squared ………………………………………….………………………22

Worked examples ……..……...……………………………….23

Questions…………………... ……………..…………………..29

Spearman’s Rank Correlation Coefficient……………..……..…………...33

Worked examples ……..…………………………………....…34

Questions……………….....………………..………………….43

Student’s t test……………..……..…………………………………..……...51

Worked examples ……..…………………………………....…52

Questions……………….....………………..………………….57

Final mixed questions…………………………….……..………..….60

2
About Statistics -

There are two main kinds of statistics.

Measuring ‘Spread’
(Descriptive statistics)
In the Biology A-level, we will use descriptive statistics to measure the ‘spread’ of data.
Descriptive statistics are used to describe data. For example, if you were investigating the
number of visitors to a beach in August (nice job if you can get it!), you might draw a graph
to see how the number of visitors varied each day, work out the average number of visitors
each day (using mean, mode or median), work out the range of visitor numbers each day.
This would all be descriptive statistics. Descriptive statistics also involve using:

 Range
 ± 2 Standard Deviations from the mean Can NOT use the
 ± 1.96 Standard Errors of the mean term “significant”

Using what we know to make inferences about what we don’t know

(Inferential statistics)
In the Biology A-level, we will use Inferential statistics which are techniques that allow us to
use what we know to make inferences (i.e. judgements) about what we don’t know. For
example, if we asked 200 people who they were going to vote for, on the day before the
local election, we could try to predict which party would win the election. We need to choose
a statistical test to use:

 Chi squared
 Spearman’s rank correlation coefficient CAN use the term
 Student’s t-test “significant”

A note about using calculators

We expect that you will often use electronic devices to calculate test statistics during your classwork.
In written examinations, you will not be asked to perform a calculation using a statistical test. It will be
important for you to understand how to select a statistical test that is appropriate for given data and to be
able interpret the results of such a statistical test. You could also be asked to explain your choices and
interpretation.

3
Measuring ‘Spread’

Distribution 1 Distribution 2

‘Distribution 1’ is tall and thin and ‘distribution 2’ is short and fat yet they have the same
mean average (and this data also has the same ‘mode’ and same ‘median’ average too.)
Nevertheless, the distributions have different spreads.

Range

Range = largest value – smallest value

To measure the spread, we could calculate the range, but sometimes there are data which
are outliers. Calculating range is simple but often not a good measure of the spread.
Using the standard deviation as a measure of spread about the mean is often a better
measure of spread.

Standard deviation
4
Using standard deviation is better than the range as it uses all the observations, and is
less affected by the outliers.

A “thin” curve means that most values remain close to the average, and the standard
deviation is small.

A “fat” curve means that there is a wider spread of values about the mean, and the
standard deviation is large.

How to calculate Standard deviation

The standard deviation is calculated using the formula:

𝚺(𝐱 − 𝐱̄ )𝟐
𝑺𝑫 = √
𝐧−𝟏

SD = Standard deviation

X = value

x̄ = mean

n = the number of values you have

5
± 2 Standard Deviations

When data has a ‘normal distribution’ as shown in the lovely ‘bell-shaped’ graph above, we
can see that just over 95% of the data is within two standard deviations either side of the
mean. We can make use of this when comparing data.

± 2 Standard deviations Worked Example 1. Lions

You have found the following ages of lions in two different zoos. The lions were randomly
selected from all the lions in each zoo.

Age of Lions at Bristol Zoo (months) Age of Lions at London Zoo (months)
36 46
31 50
35 48
24 49
21 51
47 49

6
You can use the calculators to calculate the mean and standard deviation, using the
instructions on the calculator to help!

The standard deviation is calculated using the formula:

𝚺(𝐱 − 𝐱̄ )𝟐
𝑺𝑫 = √
𝐧−𝟏

The best way to show these calculations is in do this is in a table.

Bristol Zoo London Zoo

mean 32.3 48.8
SD 9.3 1.7
2 x SD 18.6 2.4
Mean + (2 x SD) 50.9 51.2
Mean - (2 x SD) 13.7 46.4

Describing the results

We can draw a bar chart of the mean and plot the ± 2 Standard deviations from the mean
and look at the overlap of the bars.

There is an overlap in the (±2 SD) bars.

This indicates that the differences in the means (the age

of the Lions at Bristol zoo and London zoo) are likely to
be due to chance.

Note: You cannot

say how ‘likely’ this
is due to chance –
just that it is likely!
7
± 2 Standard deviations Worked Example 2.Fish

You have measured the sizes of the different genders of a species of tropical fish.

Length of female fish (cm) Length of male fish (cm)

18 8
18 10
21 10
23 12
25

Now calculate the mean and standard deviation.

Female Male
mean 21 10
SD 3 1.6
2 x SD 6 3.2
Mean + (2 x SD) 27 13
Mean - (2 x SD) 15 7

Describing the results

We can draw a bar chart of the mean and plot the ± 2 Standard deviations from the mean
and look at the overlap of the bars.

There is no overlap in the (±2 SD) bars.

This indicates that the differences in the means (the size of the
fish) is unlikely to be due to chance.

Note: You cannot say how

8
‘unlikely’ this is due to
chance – just that it is
unlikely!
± 2 Standard deviations Question 1. Heart Rate

Compare the data for resting heart rate whilst watching two different TV shows. Describe
the data.

Heart rate (beats per min) whilst watching ..

Judge Judy Judge Rinder
117 95
156 155
124 131
128 160
139 145
143 98

Now calculate the mean and standard deviation.

mean
SD
2 x SD
Mean + (2 x SD)
Mean - (2 x SD)

Describing the results

Plot the bar chart on graph paper and draw on the 2 x SD bars. You may find you can
describe the data without plotting the bar chart though.

There is an / no overlap in the (±2 SD) bars.

This indicates that the differences in the means

(………………………………..) is unlikely/likely to be due to chance.

9
± 2 Standard deviations Question 2. Drugs

Describing the results

Compare the data shown in the graph above. This figure depicts two experiments, A and B.
In each experiment, control and treatment measurements were obtained. The graph shows
the mean difference between control and treatment for each experiment. A positive number
denotes an increase; a negative number denotes a decrease. The bars show the 2 x SD for
those differences.

.
There is ……………………………………………….

This ……………………………………………………………………………………………

………………………………………………………………………………………………….

10
Standard Error of the Mean and 95% confidence
interval
The Standard Error of a mean is calculated using the following formula:

When you take a sample and calculate the mean, it is important to remember that this is
only an estimate of the true mean for the whole of the population you are measuring. If you
took a second sample, you would probably arrive at a slightly different estimate of the
mean. There is no reason to suppose that the real mean will be exactly equal to the sample
mean. It is likely to be close to it, however, and
the amount by which it is likely to differ from the
estimate can be found from the standard error.

What we do is find values (either side of the

sample mean) which are likely to include the real
mean, and say that we estimate the real mean to
lie somewhere between the upper and lower
values. If you look at the graph you can see that
there is a 95% chance that the true mean is within
1.96 x SE either side of the mean of your sample.

The 95% Confidence Interval is then calculated using the following formula:

𝟗𝟓% 𝑪𝑰 = 𝐱̄ ± 𝐒𝐄 × 𝟏. 𝟗𝟔

We can use the 95% confidence interval to state that:

 we are 95% confident that the true mean value of the population from which the
sample was taken lies between the upper and lower confidence limits

 if the intervals of two calculated means do not overlap, we are 95% confident that
these means are different.

NOTE: People find the terms ‘standard error’ and ‘standard deviation’ confusing. We use
the term‘standard deviation’ when we are talking about distributions, either of a sample or
a
population. We use the term ‘standard error’ when we are talking about an estimate
found from a sample. If we want to say how good our estimate of the mean measurement
is, we quote the standard error of the mean. If we want to say how widely scattered the
measurements are, we quote the standard deviation.
11
Standard error of the Mean and 95% Confidence Interval

Worked Example 1. The mean mass of Guinea Pigs

The owner of a Pet shop wants to work out the average mass of small guinea pigs in her
shop. She only had time to measure the mass of a few which she randomly caught.
Calculate the mean mass with 95% confidence intervals, using the sample data below.

Mass of guinea pigs (g): 19.4 21.4 22.3 22.1 20.1 23.8 24.6 19.9 19.1

Now layout the calculations (in a table helps)

mean 21.4
n 9
√n 3
SD 1.95
SE (SD ÷ √n) 0.65
1.96 x SE 1.27
Mean + 1.96 x SE 22.7
Mean - 1.96 x SE 20.1

Describing the results

The mean mass (of the sample of Guinea pigs) = 21.4g

We are 95% confident that the true mean value

(of the whole population of Guinea pigs in the pet shop)

is between 20.1g – 22.7g

12
Standard error of the mean and 95% confidence interval

Worked Example 2. Limpets

A resident on the Isle of Wight wants to know if the size of limpets is different between the
upper and middle ledges at Bembridge rocky shore. The resident collects the following
data. (Obviously the resident is not a proper scientist otherwise they would never have
presented the data in such a terrible way!) (Sizes in mm, measured using callipers:)

Middle: 43.9, 38.4, 39.4, 44.7, 40.3, 37.8, 35.6, 56.7, 47.3, 36.0, 38.7, 37.7, 41.6, 42.5,
48.3, 34.9, 21.3, 42.5, 36.2, 46.8, 41.6, 43.5, 48.6, 39.5, 50.2, 46.9, 48.7, 42.9, 37.6, 53.4

Upper: 40.3, 35.0, 36.2, 40.8, 31.2, 30.2, 27.3, 37.7, 26.5, 25.5, 33.6, 24.8, 42.3, 38.6, 30.7,
23.2, 37.3, 34.6, 32.3, 33.0, 34.6, 38.0, 28.3, 22.5, 37.0, 45.0, 32.8, 36.9, 44.2, 28.6

The best way to show these calculations is in a table.

Middle ledge Upper ledge

mean 42.12 33.63
n 30 30
√n 5.48 5.48
SD 6.74 6.07
SE (SD ÷ √n) 1.23 1.11
1.96 x SE 2.41 2.18
Mean + 1.96 x SE 44.53 35.81
Mean - 1.96 x SE 39.71 31.45

Interpreting the results

The 95% CI of the size of snails on the Middle ledge are 39.71mm to 44.53mm.

The 95% CI of the size of snails on the Upper ledge are 31.45mm to 35.81mm.

There is no overlap. Some people prefer to use graph paper to draw a bar chart with bars to
show the confidence intervals, but it is not essential to do this. The 95% CI bars do not
overlap.
13
 There is no overlap in the 95% confidence intervals of the two calculated means
(the mean size of Limpet on the Middle ledge and Limpets on the Upper ledge).

We can be 95% confident that the means are different.

Standard error of the mean and 95% confidence interval

Worked Example 2. Weight

In the UK almost a third of adults are obese. A new diet called the South Beach diet has
become popular, and the NHS wants to assess how effective it is compared to a more
traditional low calorie diet. The following data is available. What conclusions should be
drawn from the available data?

14
 % Reduction in BMI (body mass index) after 4 weeks of completion of diet
South Beach diet Traditional Low Calorie diet
2.1 2.2
2.0 1.8
1.8 1.9
2.0 2.0
1.9 1.9
2.4

There are six values for the South Beach diet so n= 6, and five values for the traditional low
calorie diet so n=5.

South Beach diet Traditional Low Calorie diet

mean 2.0 2.0
n 6 5
√n 2.45 2.24
SD 0.21 0.15
SE (SD ÷ √n) 0.08 0.07
1.96 x SE 0.16 0.14
Mean + 1.96 x SE 2.2 2.1
Mean - 1.96 x SE 1.8 1.9

Interpreting the results

The 95% confidence intervals for the South Beach diet is 1.8 to 2.2 percent.

The 95% confidence intervals for the South Beach diet is 1.9 to 2.1 percent.

There is an overlap in the 95% confidence intervals of the two calculated means (of the %
reduction in BMI for the South beach diet and the traditional low calorie diet).

We can be 95% confident that the means are not different.

15
Standard error of the mean and 95% Confidence Interval

Question 1. The mean abundance of green-winged orchid

(Use Worked example 1 to help with this question)

The counts of the green-winged orchid in a sample of ten 1x1m quadrats placed randomly
in a hay meadow are shown below:

4 2 58 1 22 5 7 17 1 4

Calculate the mean count of the orchid (per metre2) with 95% confidence intervals.

mean
n
√n
SD
SE (SD ÷ √n)
1.96 x SE
Mean + 1.96 x SE
Mean - 1.96 x SE

Describing the results

The mean count of the sample of orchids per metre2 = ……………

We are ………………………….. that the true mean value

(of the whole population of …………………………………)

is between …………………………………..

16
Standard error of the mean and 95% confidence interval
Question 2. Beans Q: What's the fastest vegetable?

A: A runner bean

Two students have each grown a row of runner beans

in the school allotment. Each day Paul has talked to
his plants to encourage them to grow taller. Simon has
not, and he thinks that Paul is a bit bonkers. Is there
any evidence to suggest that talking to the plants has
made them grow taller? (Use worked example 2 & 3 to
help with these questions.)

Height of Runner bean plants (cm)

Paul’s plants Simon’s plants
124 131
156 155
128 160
139 145
117 95
142 65
123 117
98 212
153 160
120 60

Calculating Standard error and 95% confidence interval

Paul’s plants Simon’s plants

mean
n
√n
SD
SE (SD ÷ √n)
1.96 x SE
Mean + 1.96 x SE
Mean - 1.96 x SE

Interpreting the results

There is an/no overlap in the ……………………………….. of the two calculated means (of the mean
height of Paul or Simon’s bean plants).

We can be 95% confident that the means are/are not different.

17
Standard error of the mean and 95% confidence intervals

Q3. More Beans

Q: What kind of vegetable is jealous? A: A green bean!

Ultra-competitive Paul was a bit disappointed at the result of his last experiment, so this
time Paul has decided to grow his beans using a super new fertiliser, whereas Simon just
likes to use horse-manure. Can a conclusion be made about which fertiliser is best?
Height of Runner bean plants (cm)
Paul’s plants (grown using new fertiliser) Simon’s plants (grown using horse manure)
103 160
113 145
117 120
123 125
119 126
98 212
124 123
120 129

Calculating Standard error and 95% confidence interval

Interpreting the results

There ………………………………………………………………………………………………………..

We ……………………………………………………………………………………………………………

18
More Standard error and 95% confidence interval problems.

4. Spotted knapweed is a common weed in the USA. Two methods, chemical control and
biological control, have been used to reduce the numbers of spotted knapweed plants.
Use Standard error and 95% confidence intervals to determine whether biological
control is better than biological control?

Mean number of spotted knapweed plants per m2

Chemical control Biological control

2 2

15 3

3 3

20 5

3 4

16 3

2 2

5. Two fields, A and B, were used to grow the same crop. Random samples of crop plants
from each plot were collected and their mass determined. The results are shown in the
table. Does the evidence suggest that previous use of the field affects the mass of the
crop?

Mass of crop/kg m–2

Sample
Field A - used for grazing Field B - used for same crop
cattle in previous year in previous year

1 14.5 6.4

2 16.7 9.8

3 17.4 12.9

4 17.5 16.2

5 17.5 17.1

6 17.5 17.1

7 17.5 17.1

19
6. Mayflies are insects which lay their eggs in streams and rivers. The nymphs hatch from
the eggs and live in the water for several years. Mayfly nymphs were collected by
disturbing the gravel of a stream bed. A net placed immediately downstream caught any
animals which were washed out of the gravel. Eight samples were collected from
shallow, fast-flowing parts of the stream and eight from deeper, slow-flowing parts. The
results are given in the table. What can be concluded about the differences in the
numbers of nymphs found in shallow water and deep water?

Number of Mayfly nymphs found

Sample Number
Shallow water Deep Water
1 12 16
2 14 14
3 13 14
4 18 18
5 16 21
6 16 17
7 13 16
8 14 15

7. The table shows the relative thickness of the walls of the aorta and vena cava in four
different people. Is it possible to show that the thickness of arteries is thicker than veins?

Name Thickness / µm

Artery Vein
420 180
John
490 175
James
370 165
Daniel
120 120
David

8. A general practitioner has been investigating whether the diastolic blood pressure of
men aged 20-44 differs between farm workers and firemen. For this purpose, she has
obtained a random sample of 72 firemen and 48 farm workers and calculated the mean
and standard deviations. What conclusions can be made?
Farm workers Firemen

Mean diastolic blood

88 79
pressures (mmHg)

Standard Deviation 4.5 4.2

20
Using what we know to make inferences
about what we don’t know
Inferential Statistics
You should be aware that inferential statistics are used to test a theory, known as a
hypothesis. Before we choose a statistical test we should write a null hypothesis. The table
shows how hypotheses can be turned into null hypotheses.

Hypothesis Null hypothesis

Chickens fed maize supplemented by lipid produce There is no difference between the number of male
more male offspring than those fed maize alone. and female offspring of chickens fed maize
supplemented by lipid and those fed maize alone.
There are fewer slugs in dry areas There is no difference between the number of slugs
found in wet and dry areas
Tobacco plants exhibit a higher rate of growth when Tobacco plants do not exhibit a higher rate of
planted in soil rather than peat growth when planted in soil rather than peat.

The null hypothesis

can be thought of as
“life is dull and
nothing ever
happens” hypothesis!

Once the null hypothesis is stated, a statistical test is then chosen. This either supports or
fails to support the null hypothesis.

21
Chi-squared test
All chi-squared tests are concerned with counts of things (frequencies) that you can put into
categories. For example, you might be investigating flower colour and have counted the
numbers (frequencies) of red flowers and white flowers (categories). Or you might be
investigating human health and have frequencies of smokers and non-smokers.

The test looks at the frequencies you obtained when you counted them and compares them
with the frequencies you might expect to get in order to determine whether the difference is
significant or not.

22
Chi-squared Worked Example 1. Snails on the Seashore

Why did the periwinkle blush?

Answer: because the sea weed!!

You have been wandering about on a seashore and you have noticed that a small snail (the
flat periwinkle) seems to live only on certain types of seaweed. You decide to investigate
whether the animals prefer to certain types of seaweed by counting numbers of animals on
the different types of seaweed. You end up with the following data:

Type of Seaweed Observed frequency

(the numbers of periwinkle)

serrated wrack 45
bladder wrack 38
egg wrack 10
spiral wrack 5
other seaweed 2
TOTAL 100

Null hypothesis

The null hypothesis when doing Chi-squared is

“there is no significant difference between the observed and expected frequencies.”

In other words, the periwinkle does not have a preference for which seaweed it lives on.
This is now used to work out the ‘expected’ frequencies.

23
Expected Frequencies

Our null hypothesis is that there is no difference between the observed and expected
frequencies. If this were exactly the case there would be no differences in the frequencies
over all of our categories (i.e. the five types of seaweed). The best estimate we could make
therefore would be to add up all our observed frequencies and divide by the number of
categories. So our expected frequency for each category would be:

45 + 38 + 10 + 5 + 2 = 100

100 ÷ 5 = 20 = expected frequency

Type of Seaweed Observed frequency Expected frequency

serrated wrack 45 20
bladder wrack 38 20
egg wrack 10 20
spiral wrack 5 20
other seaweed 2 20
TOTAL 100 100

Calculating the Chi-squared value

Next we calculate the value of Chi-squared using the formula below.

O = Observed frequency

E = Expected frequency

The best way to show these calculations is in do this is in a table.

Type of Observed Expected (O-E)2

O-E (O-E)2
Seaweed frequency frequency E
serrated wrack 45 20 25 625 31
bladder wrack 38 20 18 324 16
egg wrack 10 20 -10 100 5
spiral wrack 5 20 -15 225 11
other seaweed 2 20 -18 324 16
Total = 79

The total of our final column is the Chi-squared value. ( = 79)

24
The ‘critical value’ & the ‘degrees of freedom’

Before we can interpret our results we need to work out the ‘critical value’. The critical value
represents the borderline between accepting or rejecting our null hypothesis. We get the
critical value from the data sheet, but this depends on the number of ‘degrees of freedom’.

Degrees of freedom = number of categories -1

'Degrees of freedom' is a term that can be bit confusing. A simple (though not completely
accurate) way of thinking about degrees of freedom is to imagine you are picking people to
play in a team. You have eleven positions to fill and eleven people to put into those
positions. How many decisions do you have? In fact you have ten, because when you come
to the eleventh person, there is only one person and one position, so you have no choice.
You thus have ten 'degrees of freedom' as it is called. So 11 categories but only 10
‘degrees of freedom’. Hence, degrees of freedom = number of categories -1

Likewise, the periwinkle snail was found on the serrated wrack, bladder wrack, egg wrack,
spiral wrack, or other seaweed. There are five categories (five different types of seaweed),
so only 4 degrees of freedom.

Interpreting the results

Chi-squared gives a number which indicates how big the difference is between the
observed data and the expected data.

If the Chi-squared value is small, then there is a small difference between the observed and
the expected data. This means the null hypothesis is accepted (likely to be correct). In other
words, the snails don’t mind which seaweed they live on!

If the Chi-squared value is huge, then there is a huge difference between the observed and
the expected data. This means the null hypothesis is rejected. In other words, the snails do
indeed have a preference for living on a particular seaweed.

Looking at the table above we can see that the critical value of Chi-squared at 5%
significance (p=0.05) and 4 degrees of freedom is 9.49.

25
Our calculated value is 79

The calculated value is bigger (much bigger!) than the critical value. In a chi-squared test
this means we must reject the null hypothesis. In doing this we are saying that the snails
are not scattered about the various sorts of seaweed randomly. Biologists would infer that
this means they seem to prefer living on certain species.

Our calculated value of Chi-squared is much larger than the critical

value of Chi-squared.

There is less than 5% probability that the differences (between the

observed and expected data) are due to chance.

We reject our null hypothesis.

It is worth pointing out that statistics of this kind tell you nothing about the biology of the
situation. All we are saying is that our observed frequencies are different to our expected
ones. (For example, you could criticise our approach by pointing out that it might be that
there are not equal amounts of each type of seaweed on the shore for the animals to live
on.)

26
Chi-squared Worked Example 2. Birds on the Bird-Table

Q: When should you buy a bird?

A: When it’s going cheep!

Three neighbours have very similar bird-tables in their gardens. Bill owns the middle garden
and is a keen birdwatcher but he suspects that Kate, one of his neighbours, is actively
encouraging birds away from Bill’s garden into her own somehow. Is Bill right?

Observed frequency
Garden
(the numbers of bird’s visiting the
garden on one day in March)
Bill 112
Kate 145
Chris 139
TOTAL 396

Null hypothesis

The null hypothesis when doing Chi-squared is

“there is no significant difference between the observed and expected frequencies.”

Expected Frequencies and Calculating the Chi-squared value

We would expect the same numbers of birds to visit each garden if the null hypothesis is
correct. A total of 396 birds were seen, so we would expect 132 in each garden (396 ÷ 3)

Observed Expected (O-E)2

Garden O-E (O-E)2
frequency frequency E
Bill 112 132 -20 400 3.0
Kate 145 132 13 169 1.3
Chris 139 132 7 49 0.4
= 4.7

27
Degrees of Freedom

Degrees of freedom = number of categories -1

We have three categories (i.e. three gardens) so the degrees of freedom is 3-1 = 2

Interpreting the results

Looking at the table above we can see that the critical value of Chi-squared at 5%
significance (p=0.05) and 2 degrees of freedom is 5.99.

Our calculated value is 4.7

The calculated value is smaller than the critical value. In a chi-squared test this means we
must accept the null hypothesis. (In other words, Bill is not right! – Kate is not secretly
putting out expensive bird-food to encourage the birds into her garden!)

Our calculated value of Chi-squared is smaller than the critical value

of Chi-squared.

There is more than 5% probability that the differences (between the

observed and expected data) are due to chance.

We accept our null hypothesis.

28
Chi-squared Question 1. Mendel and his Peas

Q: What do you call an angry pea?

A: Grump-pea.

Mendel planted some round peas which grew into plants that produced a total of 556 peas.
423 were round peas and 133 were wrinkled peas. Mendel postulated that round is
dominant to wrinkled, and he work out the expected ratio for two heterozygous parent
plants as 3:1. Do his experimental data support his 3:1 expected ratio?

Null hypothesis Write the null hypothesis here.

…………………………………………………………………………………………………………..

…………………………………………………………………………………………………………..

Calculating the Chi-squared value

Observed Expected (O-E)2

Shape of pea O-E (O-E)2
frequency frequency E

Interpreting the results

How many degrees of freedom are there? ………………

What is the critical value (from the table)? …………………..

Our calculated value of Chi-squared is larger/smaller than the critical

value of Chi-squared.

There is more/less than 5% probability that the differences

(between the observed and expected data) are due to chance.

We accept/reject our null hypothesis.

29
Chi-squared Question 2. A zookeeper’s dilemma

Q: What do you call a flying primate? A: A hot air baboon!

A zookeeper thinks that lowering the intensity of the light in the

primate exhibits will reduce the amount of aggression between the
baboons. In exhibit A, with a lower light intensity, he observes 36
incidences of aggression over a one month period. In exhibit B,
with normal lights, he observes 42 incidences of aggression. Does
he have enough evidence to support his theory?

Null hypothesis Write the null hypothesis here.

…………………………………………………………………………………………………………..

…………………………………………………………………………………………………………..

Calculating the Chi-squared value

Interpreting the results

How many degrees of freedom are there? ………………

What is the critical value (from the table)? …………………..

Our calculated value of Chi-squared is larger/smaller than the critical

value of Chi-squared.

There is more/less than 5% probability that the differences

(between the observed and expected data) are due to chance.

We accept/reject our null hypothesis.

30
Chi-squared Question 3. The West-Africian bee-eaters
You have just returned from a 3 year stint in the jungles of western
Africa, where you studied the habitat selected by the native bee-eaters
(a family of birds that specialize in catching bees and wasps on the
wing, taking them to a perch, bashing their stingers out, and devouring
them. In a pinch, they will eat other flying or hopping insects, such as
grasshoppers). Several habitats were available to the bee-eaters. Is there evidence to
suggest that birds prefer a particular habitat?

Forest Understory Canopy Emergent Grassland Field River-

floor layer layer layer bank
Number
3 15 17 20 3 11 4
of birds

Null hypothesis Write the null hypothesis here.

…………………………………………………………………………………………………………..

…………………………………………………………………………………………………………..

Calculating the Chi-squared value

Interpreting the results

How many degrees of freedom are there? ………………

What is the critical value (from the table)? …………………..

Our calculated value of Chi-squared is larger/smaller than the critical

value of Chi-squared.

There is more/less than 5% probability that the differences

(between the observed and expected data) are due to chance.

We accept/reject our null hypothesis.

31
More Chi-squared problems.

4. One section of a river was trawled and four species of fish counted and frequencies
recorded. There were 15 Rudd, 15 Roach, 4 Dace and 6 Bream. Are the fish present
in the river in equal proportions?

5. An optician noticed the following information about colour-blindness in males and

females. Is there a significant difference between the between the observed
frequency of colour blindness in males and females?

Observed frequencies Males Females

Colour blind 56 14
Not colour blind 754 536

6. The table below shows the number of patients requesting an urgent appointment to
see a Doctor on particular days of the week. Are the differences significant?

Monday Tuesday Wednesday Thursday Friday

Numbers
125 88 87 94 108
of patients

7. Cranes are large birds. Biologists have used DNA hybridisation to confirm the
relationships between different species of crane. They made samples of hybrid DNA
from the same and from different species. They measured the percentage of
hybridisation of each sample. The results are shown in the table. Are there any
differences which are statistically significant?

Species of crane Mean percentage DNA hybridisation

Grus americana and Grus monachus 97.4

Grus monachus and Grus rubicunda 95.7

Grus americana and Grus rubicunda 95.5

Grus rubicunda and Grus rubicunda 99.9

Grus americana and Grus americana 99.9

Grus monachus and Grus monachus 99.8

8. The table shows the number of cases of tuberculosis in the East Midlands between
2000 and 2005. Are the differences in number of cases of tuberculosis significant?

Number of cases of 2000 2001 2002 2003 2004

TB per 100,000 of
the population 10.6 11.1 11.9 7.9 9.9

32
Spearman’s Rank

Spearman’s rank correlation is a statistical test that is carried out in order to assess the degree of
association between different measurements from the same sample. That is, if you are looking for a
positive or negative correlation between two variables.

Positive correlation No correlation Negative correlation

When the points on a graph clearly fit onto a line of best fit it is easy to determine whether a correlation
exists. However, as the points become further placed from each other it is hard to make an accurate
judgement. This is where statistics is used; to clarify how confident we are that a correlation exists.

33
Spearman’s Rank Worked Example 1. Whales

Q: What do you get if you

cross an elephant with a
whale?

A: A submarine with a built-

in snorkel.

Do blue whales get heavier as they get longer? From the table below it certainly looks as if they do. If
we drew a graph we would get an ‘uphill line’ (a positive correlation). However, to find out if the
correlation is statistically significant we must calculate Spearman’s rank correlation coefficient ( rs ).

Length (metres) Mass (tonnes)

1 1.5
1.5 2.3
2.3 3.6
3.4 7.1
4.4 2.6
4.6 13.3
6.2 7.5
7 11.2
7 12.1
8.7 11
10.5 12
12 18.2

Null hypothesis

When doing Spearman’s rank the two variables are used to construction the null
hypothesis. The null hypothesis always assumes there is no relationship.

“There is no correlation between (variable 1) and (variable 2)”

In other words, for this example:

“There is no correlation between the length of a blue whale and the mass of the whale.”

34
Rank the data

For each set of data assign ranks from lowest to highest. The lowest value in a column will be
given the rank of 1, the next smallest number will be given a 2 etc. If there are tied scores each of those
will share the ranks and be given the average (mean) rank.
For example, there are two whales with the same length (7 metres). These would be ranked 8 and 9, so
they are given the mean value. (i.e. 8+9 = 17 17÷2 = 8.5).

Length metres) Rank of Mass (tonnes) Rank of

(variable 1) variable 1 (variable 2) variable 2

1 1 1.5 1
1.5 2 2.3 2
2.3 3 3.6 4
3.4 4 7.1 5
4.4 5 2.6 3
4.6 6 13.3 11
6.2 7 7.5 6
7 8.5 11.2 8
7 8.5 12.1 10
8.7 10 11 7
10.5 11 12 9
12 12 18.2 12

Calculate Spearman’s rank

Next we have to work out D2 ;this is the difference in the rankings, squared.

35
Then we calculate the value of Spearman’s rank correlation, rs, using the equation below.

D = difference between the rank of the paired measurements

n = number of paired measurements

∑ = the sum of

Add up the D2 column (∑D2). ∑D2 = 47.5

If we substitute the numbers into the equation we get:

= 1- ((6 X 47.5) ÷ (123-12)

= 1 - (285 ÷1716)

= 1 – 0.166

36
 rs = 0.834

Interpret the results

The first thing we notice is that the answer is a positive number, so we know that length are
positively correlated (i.e. not negatively correlated.)

The closer rS is to 1 or -1 the more likely the correlation. A perfect positive correlation has
an rS value of 1, a perfect negative correlation has a value of -1.

If the value lies between -1 and 1 we need to carry out a test for significance.

37
Before we can interpret our results we need to work out the ‘critical value’. The critical value
represents the borderline between accepting or rejecting our null hypothesis.

We have 12 paired values which gives us a critical value of 0.59 (for a positive correlation)
or -0.59 (for a negative correlation). This means that any value between +0.59 and +1 is a
statistically significant (positive) correlation.

The Spearman’s rank correlation coefficient, rs, between the length and mass of a blue
whale was +0.834. The calculated value is bigger than the critical value. In a Spearman’s
rank test this means we must reject the null hypothesis. In doing this we are saying that
there is a relationship (a positive correlation) between the length and mass of the blue
whale.

Our calculated value of Spearman’s rank correlation

coefficient, rs (+0.834) is larger than the critical
value of (+0.59).

There is less than 5% probability that the positive

correlation between the length and mass of the
whale is due to chance.

We reject our null hypothesis.

38
Spearman’s Rank Worked Example 2. Heather in a Moorland

Q: What did the big flower say to the

small flower?
A: What's up, bud?

During monitoring of a moor, an ecologist collected data from an area of moorland that had been
restored in 2003. In order to assess whether a correlation existed between the two plant species she
studied or whether they were growing independently of one another, she used a quadrat to collect the
following data.

Number of
Number of
Common
Bilberry plants
Heather plants
(per m2)
(per m2)

50 18
175 12
270 20
375 10
425 10
580 12
710 8
790 6
890 10
980 9

Null hypothesis

“There is no correlation between the number of Bilberry plants growing in an area and
the number of Common Heather plants growing in the area”

39
Rank the data & Calculate Spearman’s rank

Number of
Number of
Common Rank Common
Bilberry plants Rank Bilberry
Heather Heather Difference (D) D2
(per m2) (variable 1)
plants (variable 2)
(per m2)
50 1 18 9 -8 64
175 2 12 7.5 -5.5 30.25
270 3 20 10 -7 49
375 4 10 5 -1 1
425 5 10 5 0 0
580 6 12 7.5 -1.5 2.25
710 7 8 2 5 25
790 8 6 1 7 49
890 9 10 5 4 16
980 10 9 3 7 49
2
∑D = 285.5

∑D2 = 285.5

n = 10 (because we have 10 paired readings for the plants)

If we substitute the numbers into the equation we get:

= 1- ((6 X 285.5) ÷ (103-10)

= 1 - (1713 ÷990)

= 1 – 1.730

rs = - 0.730

40
Interpret the results

The answer is a negative number, so we know that relationship between the two plants
shows a negative correlation. We have 10 paired values which gives us a critical value
(from the table of critical values) of -0.65 (for a negative correlation).

Our calculated value of Spearman’s rank correlation

coefficient (-0.730) is larger than the critical value
of (-0.65).

There is less than 5% probability that the negative

correlation between the Bilberry and Common
Heather is due to chance.

We reject our null hypothesis.

Spearman’s Rank Worked Example 3. Frigate birds

Males of the Frigatebird have a large red throat pouch. They visually display this pouch and
use it to make a drumming sound when seeking mates. Researchers wanted to know
whether females, who presumably choose mates based on their pouch size, could use the
pitch of the drumming sound as an indicator of pouch size. They estimated the volume of
the pouch and the frequency of the drumming sound in 16 males:

41
 Volume
Frequency (Hz)
(cm3)
1760 529
2040 390
2440 473
2550 461
2730 465
2740 532
3010 484
3080 527
3370 488
3740 485
4910 478
5090 434
5380 468
5850 449
6730 464
6990 530

Null hypothesis

“There is no correlation between the volume of the bird’s pouch and frequency of the
sound it makes.”

Rank the data & Calculate Spearman’s rank

Volume Rank
Rank Volume Frequency
(per cm3) Frequency Difference (D) D2
(variable 1) (Hz)
(variable 2)
1760 1 529 15 -14 196
2040 2 390 1 1 1
2440 3 473 9 -6 36
2550 4 461 5 -1 1
2730 5 465 7 -2 4
2740 6 532 17 -11 121
3010 7 484 11 -4 16
3080 8 527 14 -6 36
3370 9 488 13 -4 16
3740 10 485 12 -2 4
4910 11 478 10 1 1
5090 12 434 3 9 81
5380 13 468 8 5 25
5850 14 449 4 10 100
6730 15 464 6 9 81
6990 16 530 16 0 0
2
∑D = 719

42
∑D2 = 719

n = 16

If we substitute the numbers into the equation we get:

= 1- ((6 X 719) ÷ (163-16)

= 1 - (4314 ÷ 4080)

= 1 – 1.057

rs = - 0.057

Interpret the results

The answer is a negative number, so we know that any relationship between volume and
frequency shows a negative correlation. We have 16 paired values which gives us a critical
value (from the table of critical values) of -0.51 (for a negative correlation).

Our calculated value of Spearman’s rank correlation coefficient (-0.057) is

smaller than the critical value of (-0.51).

There is more than 5% probability that the correlation between the volume
of the pouch and the frequency of the sound it makes is due to chance.

We accept our null hypothesis.

43
Spearman’s rank Question 1. Purple loosestrife control

Q: What do you call a beetle that can't have too much sugar?

A: a diabeetle.

A European beetle was tested to see whether it could be used for the biological control of
purple loosestrife in the USA. In an investigation, beetles were released in an area where
purple loosestrife was a pest. The table shows some of the results. Is it possible to prove
that the beetles are effective in controlling purple loosestrife?

Mean number of Purple loosestrife plants (per m2) Mean number of Beetles (per m2)
28 4
22 5
8 40
6 62
7 68

Null hypothesis Write the null hypothesis here.

…………………………………………………………………………………………………………..

…………………………………………………………………………………………………………..

Rank the data & Calculate Spearman’s rank

Mean number
Mean number
purple Rank Purple
Beetles Rank Beetles
loosestrife loosestrife Difference (D) D2
(per m2) (variable 2)
(per m2) (variable 1)

∑D2 =

44
∑D2 = ………………….

n = ……………………...

Substitute the numbers into the equation to calculate rs

rs = ………………

Interpret the results

Our calculated value of Spearman’s rank correlation coefficient is

smaller/bigger than the critical value.

There is less/more than 5% probability that the (positive/negative)

correlation between …………………………………………….……………….……………………
and ……………………………………………….….………………… is due to chance.

We accept/reject our null hypothesis.

45
Spearman’s rank Question 2. Blood sugar control

Q: Did you hear the joke about the peanut butter?

A: I'm not telling you. You might spread it!

A student ate a meal containing carbohydrates at 07:00. He ate nothing else for the next
five hours. The table shows the concentration of glucose in his blood at hourly intervals
after the meal. Is there a significant relationship between time of day and blood sugar
concentration?

Time of day Concentration of glucose in blood

(mg per 100 cm3 of blood)
07:00 90
08:00 120
09:00 70
10:00 85
11:00 110
12:00 80

Null hypothesis Write the null hypothesis here.

…………………………………………………………………………………………………………..

…………………………………………………………………………………………………………..

Rank the data & Calculate Spearman’s rank

∑D2 =

46
∑D2 = ………………….

n = ……………………...

Substitute the numbers into the equation to calculate rs

rs = ………………

Interpret the results

Our calculated value of Spearman’s rank correlation coefficient is

smaller/bigger than the critical value.

There is less/more than 5% probability that the (positive/negative)

correlation between …………………………………………….……………….……………………
and ……………………………………………….….………………… is due to chance.

We accept/reject our null hypothesis.

47
Spearman’s rank Question 3. Biodiversity on roundabouts

Roundabouts are common at road junctions in towns and cities. Ecologists investigated the
species of plants and animals found on roundabouts in a small town. The grass on the
roundabouts was mown at different time intervals. The table shows the mean number of
plant species found on the roundabouts. From the data, is it possible to prove that mowing
too frequently reduces biodiversity?

Approximate interval between mowing Mean number of plant species

(days)
7 15.8
14 8.3
40 21.2
80 30.6
365 32.0

Null hypothesis Write the null hypothesis here.

…………………………………………………………………………………………………………..

…………………………………………………………………………………………………………..

Rank the data & Calculate Spearman’s rank

∑D2 =

48
∑D2 = ………………….

n = ……………………...

Substitute the numbers into the equation to calculate rs

rs = ………………

Interpret the results

Our calculated value of Spearman’s rank correlation coefficient is

smaller/bigger than the critical value.

There is less/more than 5% probability that the (positive/negative)

correlation between …………………………………………….……………….……………………
and ……………………………………………….….………………… is due to chance.

We accept/reject our null hypothesis.

49
More Spearman’s rank correlation coefficient problems.

4. A student set up an experiment as shown opposite.

The water bath was heated to 30⁰C, and the
yeast left for 5 mins to allow the temperature of
the yeast to equilibrate with the temperature of the
water bath.
The rate of respiration in the yeast was then
measured by recording the number of bubbles
produced over 1 minute.
The experiment was then repeated at 40, 50, 60,
70, 80 and 90⁰C.

The final results are shown below

Temperature (⁰C) Rate of respiration

(Bubbles per minute)
10 2
20 5
30 9
40 17
50 32
60 12
70 4
80 0
90 0

What, if any, is the relationship between temperature and rate of respiration?

Use Spearman’s Rank Correlation Coefficient (rs) in your analysis of the results.

5. Great tits are small birds. In a study of growth in great tits, the relationship between
the mass of the eggs and the mass of the young bird on hatching was investigated.
Is there a relationship?

Egg mass / g Chick mass / g

1.37 0.99
1.49 0.99
1.56 1.18
1.70 1.16
1.72 1.17
1.79 1.27
1.93 1.75

50
51
6. A student carried out an investigation to find out if there is a link between Mussel
shell length and width on a rocky shore. Is there a relationship?

Shell length (mm) Shell width (mm)

46 23
group A / mm, x1 group B / mm, x2
50 28
45 41
45 31
63 26
57 33
65 35
73 21
55 38
79 30
62 36
59 38
71 45
68 28
77 42

7. In a study of the 18 volunteers, the correlation between the mood and the amount of
liquid consumed by daily drinking was investigated. The Spearman’s rank
correlation, rs = 0.12 was obtained. How should this data be interpreted?

8. The correlation value obtained in a study of correlation between body height and
biological age was rs = 0.97. May we conclude that height and age are definitely
excellently correlated?

9. A student carried out 6 samples examining the growth rate of bacteria and the
concentration of citric acid present in the growth medium, and then calculated an r s
value of -0.89. How should this data be interpreted?

52
Student’s t-test
Use this test when you are looking for the difference between two means, and you want to
know if the difference is ‘significant’ or not. OK, so the formula looks a bit scary, but if you
look through the worked examples you’ll realise it’s not so tough!

53
Student’s t-test Worked Example 1. Bacteria

We have been growing two different strains of bacteria in flasks containing glucose. We had
4 replicate flasks for each bacterium. We have measured the biomass and want to find out
whether or not the results are significantly different for the two different strains of bacteria.

Mass (milligrams) of bacteria

Bacterium A Bacterium B
Flask 1 520 230
Flask 2 460 270
Flask 3 500 250
Flask 4 470 280
Mean value 487.5 257.5

Null hypothesis

The null hypothesis when doing the Student t-test is

“there is no significant difference between the two different means”

54
Calculating the value of t

Next we calculate the value of t using the formula below.

The best way to show these calculations is in a table.

Bacterium A Bacterium B
Mean 487.5 257.5
N 4 4
s (standard deviation) 27.54 22.17
s2 758.5 491.5
s2 ÷ n 189.6 122.9

Substitute these values into the formula

487.5 – 257.5
t = √(189.6 +122.9)

230
t = √312.5

230
t = √17.7

t = 12.99

55
The ‘critical value’ & the ‘degrees of freedom’

Before we can interpret our results we need to work out the ‘critical value’. You will
remember from Chi-squared that this represents the borderline between accepting or
rejecting our null hypothesis. We get the critical value from the data sheet, but this depends
on the number of ‘degrees of freedom’. Hopefully you will remember all about 'degrees of
freedom' from Chi-squared. The calculation is slightly different simply because it allows for
two sets of data.

There were 4 flasks for each of the bacteria (n=4 for both bacteria)

Hence:

The number of degrees of freedom = (4 + 4) – 2

The number of degrees of freedom = 6

So we can see from the table of critical values of t, that 6 degrees of freedom = 2.48

Our value of t = 12.99 This is much higher than the critical value.

Interpreting the results

Our calculated value of t is greater than the critical value of t.

There is more than 5% probability that the differences in the means (mean
mass of bacterium A and mean mass of bacterium B) are not due to chance.

We reject our null56

hypothesis.
Student’s t-test Worked Example 2. Enzymes

Some species of bacteria cause diseases of the stomach. Most are killed by acid gastric
juices produced by the stomach lining. Some species of bacteria survive the antibacterial
action of gastric juices by secreting the enzyme urease. This enzyme catalyses a reaction
that produces ammonia. The ammonia neutralises the acid in gastric juice. A student
believes that a small increase in temperature reduces the effect of urease and has
produced the table of results below. The student wants to know whether her findings are
significant or not.

Time for the acid to be neutralised (s)

36.5oC 37.5oC
57 58
43 57
49 51
51 57
44 54
No data 49

Null hypothesis

The null hypothesis when doing the Student t-test is

“there is no significant difference between the two different means”

Calculating the value of t

36.5oC 37.5oC
Mean 48.6 54.3
N 5 6
s (standard deviation) 5.68 3.67
s2 32.26 13.47
s2 ÷ n 6.5 2.2

57
Substitute these values into the formula

48.6 – 54.3
t=
√(6.5 + 2.2)

5.7
t = √8.7 (Ignore the minus sign, as the ‘difference in means’ is intended)

5.7
t = 2.95

t = 1.9

The ‘critical value’ & the ‘degrees of freedom’

There were n = 5 for the lower temperature, but n = 6 for the higher temperature

Hence:

The number of degrees of freedom = (5 + 6) – 2

The number of degrees of freedom = 9

So we can see from the table of critical values of t, that 6 degrees of freedom = 2.26

Our value of t = 1.9 This is lower than the critical value.

58
Interpreting the results

Our calculated value of t is less than the critical value of t.

There is more than 5% probability that the differences in the means (mean
mass of bacterium A and mean mass of bacterium B) are due to chance.

We accept our null hypothesis.

Student’s t-test Question 1.

A theatre nurse suspects that his newly purchased bit of expensive kit that gives instant
blood - haemoglobin levels is faulty. So he takes readings and compares these with his
trusty old bit of kit. His results are in the table below. Are his results significantly different?

Haemoglobin reading (g/dL)

New expensive kit Old bit of kit
17.6 16.5
13.9 12.5
15.4 14.8
16.5 15.2
13.5 12.1
16.5 15.3

59
The Null hypothesis is

…………………………………………………………………………………………………………

…………………………………………………………………………………………………………

Calculating the value of t

New expensive kit Old bit of kit

Mean
N
s (standard deviation)
s2
s2 ÷ n

Substitute these values into the formula

−
t=
√( + )

t=
√

t=

t=

The ‘critical value’ & the ‘degrees of freedom’

Now calculate how many Degrees of freedom

The number of degrees of freedom = ……………………………

Use the table on the previous page to find the critical values of t = ……………………

Interpreting the results

Our calculated value of t is less/greater than the critical value of t.

There is more than 5% probability that the differences in the means (mean
mass of bacterium A and mean mass of bacterium B) are/not due to chance.

We accept/reject our null hypothesis.

60
Student’s t-test Question 2.
A scientist is examining the rate of mitosis in the root cells and shoot cells of a species of
grass. She wants to know whether or not the rate of cells division in the root is quicker than
the rate of mitosis in the shoot cells. Are her results significantly different?

Time for one cell cycle (hours)

Shoot cells Root cells
2.6 2.1
3.5 1.7
4.1 2.6
2.8 3.8
2.7 3.5
2.5 1.9
3.1 2.1
2.9 2.5
3.4 2.2

The Null hypothesis is

…………………………………………………………………………………………………………

…………………………………………………………………………………………………………

Calculating the value of t

New expensive kit Old bit of kit

Mean
N
s (standard deviation)
s2
s2 ÷ n

Calculate the value of t

The number of degrees of freedom = ……………………………

Use the table on the previous page to find the critical values of t = ……………………

Interpreting the results

Our calculated value of t is …………………………………………………………………….

There is ……………………………………………………………………………………………………………………………………………………………………..
61
We …………………………………………………………………………………………………………..
Final Questions.
For each of the following questions, use the flow diagram below to help you to use
the most appropriate statistical test.

Flowchart for deciding which Inferential statistical test to use

Decide if you are looking for:

associations between data
or
differences between samples

Yes
Looking for Correlation
association? coefficient

No

Yes
Comparing Chi-squared
frequencies? test

No

Student’s t-test

62
1. The two-spot ladybird is a small beetle. It has a red form and a black form. These two
forms are shown in the diagram.

Colour is controlled by a single gene with two alleles. The allele for black, B, is dominant
to the allele for red, b. Scientists working in Germany compared the number of red and
black ladybirds over a six-year period. They collected random samples of ladybirds from
birch trees. Some of the results from the investigation are shown in the table.

How could you show that the frequency of the allele has remained the same?

Which statistical test should you use? Justify your choice of statistical test.

State the null hypothesis and interpret the results using the terms probability &
chance.

Year Season Frequency of b allele

1933 Autumn 0.70

1934 Autumn 0.82

1935 Autumn 0.59

1936 Autumn 0.76

1937 Autumn 0.57

1938 Autumn 0.78

63
2. Fur seals live in Antarctic seas. They feed on fish and shrimp-like animals called krill.
During the summer the fur seals come ashore to breed. The table shows the number of
fur seals breeding on an Antarctic island from 1956 to 1986.

How could the increase in adult fur seal numbers be shown to be significant?

Which statistical test should you use? Justify your choice of statistical test.

State the null hypothesis and interpret the results using the terms probability &
chance.

Year Number of adult fur seals

1956 100

1964 100

1970 200

1975 100

1976 1600

1981 2900

1983 3100

1986 11700

3. A young bird watcher was watching a pair of breeding blue tits bringing food back to the
nest for the newly hatched chicks. He measured their ‘return rate’ which is a factor that
takes into account the bird’s time away from the nest, and the success in returning to the
nest with food for the chicks. He created a table of his results. Is the male blue tit a
significantly better provider for the chicks than the female?
Gender Return rate (mg/hour)
Female
adult 56 75 45 71 61 64 58 80 76 61

Male
adult 66 70 40 60 65 56 59 77 67 63

Which statistical test should you use? Justify your choice of statistical test.

State the null hypothesis and interpret the results using the terms probability &
chance.

64
4. Malaria is a disease caused by a parasite. Scientists investigated the effect of malaria
on competition between two species of Anolis lizard on a small Caribbean island. They
sampled both populations by collecting lizards from a large number of sites on the
island.

The scientists investigated the percentage of lizards of both species that were
infected with malaria at different sites on the island. They collected samples of both
lizards at intervals of 3 months for 1 year. They also recorded the elevation (height
above sea level) of each site. Some of their results are shown in the table.

Elevation
Total number Total number
of Percentage of Percentage of
of of
collectio A. gingivinus A. wattsi
Site A. gingivinus A. wattsi
n infected with infected with
collected in collected in
site / malaria malaria
one year one year
metres

1 10 13 0 0 0

2 80 30 0 0 0

3 120 35 23 3 0

4 200 40 30 7 0

5 300 52 46 12 0

6 315 35 31 13 1

7 370 155 37 79 2

8 414 124 44 68 4

(a) A preliminary study suggested that malarial infections in A.gingivinus were more
common at higher elevations. Use the data provided to determine whether this
suggestion is statistically significant.
Which statistical test should you use? Justify your choice of statistical test.

State the null hypothesis and interpret the results using the terms probability & chance.

(b) The scientists carried out a statistical test to determine whether the correlation
between the number of A. wattsi collected and the percentage of A. gingivinus
infected was significant.
Which statistical test should you use? Justify your choice of statistical test.

State the null hypothesis and interpret the results using the terms probability & chance.

65
5. In an investigation by a student into the responses of maggots, the bottom of a large box
was marked with six coloured segments, as shown in the diagram.

30 maggots were placed on each segment in the box. A transparent cover was put on the
box and light bulbs were positioned so that the segments were evenly illuminated. The
positions of the maggots were recorded after one hour. The intensity of the light reflected by
each segment was measured. The experiment was repeated three more times. The total
number of maggots in each segment from the four experiments is shown in the table.

Colour of Intensity of reflected light /

Total number of maggots
segment arbitrary units

Black 4 154

Red 25 229

Blue 10 178

White 44 47

Green 25 48

Yellow 40 64

Give one conclusion about the responses of maggots which is supported by these results,
and test your conclusion to see if it is statistically significant, using a suitable statistical test.

Which statistical test should you use? Justify your choice of statistical test.

State the null hypothesis and interpret the results using the terms probability &
chance.

66
6. Here are the results of an investigation into the rate of photosynthesis in the pond weed
Elodea. The number of bubbles given off in one minute was counted under different light
intensities, and each measurement was repeated 5 times. How can you show that mean
rate of photosynthesis at each light intensity is significantly different?

light Rate of photosynthesis (number of bubbles/min)

intensity repeat 1 repeat 2 repeat 3 repeat 4 repeat 5
(Lux)
0 5 2 0 2 1
500 12 4 5 8 7
1000 7 20 18 14 24
2000 42 25 31 14 38
3500 45 40 36 50 28
5000 65 54 72 58 36

7. In a test of two drugs 8 patients were given one drug and 8 patients another drug. The
number of hours of relief from symptoms was measured with the following results:
Drug Time spent symptom free (hours)
A 3.2 1.6 5.7 2.8 5.5 1.2 6.1 2.9
B 3.8 1.0 8.4 3.6 5.0 3.5 7.3 4.8

Find out which drug is better by using an appropriate statistical test to find if it is
significantly better than the other drug.

8. In one of Mendel's dihybrid crosses, the following types and numbers of pea plants were
recorded in the F2 generation:
Number of Number of Yellow Number of Number of
Yellow round wrinkled seeds Green round Green wrinkled
seeds seeds seeds
395 122 96 39

According to theory these should be in the ratio of 9:3:3:1.

Use the table of critical values below to determine whether these observed results agree
with the expected ratio at P = 0.05 and P = 0.001?

67
9. The areas of moss growing on the north and south sides of a group of trees were
compared.
Orientation Total area of moss growing (m2)
North side of tree 20 43 53 86 70 54
South side of tree 63 11 21 54 9 74

Is there a significant difference between the north and south sides?

10. The table below shows the results of an experiment. Five different trays of seedlings
were grown under red or yellow light over a four-hour period. The growth of the
seedlings was measured. Are the differences in growth significant?

Mean increase in length/mm

Tray of seedlings Grown in the red light Grown in yellow light

P 5.2 3.8

Q 3.9 4.2

R 4.9 3.4

S 4.1 3.3

T 4.9 3.7

68