CAIE FP2 2012 November — Question 9 14 marks

Exam BoardCAIE
ModuleFP2 (Further Pure Mathematics 2)
Year2012
SessionNovember
Marks14
PaperDownload PDF ↗
Mark schemeDownload PDF ↗
TopicT-tests (unknown variance)
TypeTwo-sample confidence interval difference of means
DifficultyStandard +0.8 This is a standard two-sample t-test with small samples requiring calculation of summary statistics from raw data, construction of a confidence interval, and a one-tailed hypothesis test. While it involves multiple steps and careful handling of degrees of freedom, it follows a well-established procedure taught in Further Statistics modules without requiring novel insight or complex problem-solving beyond textbook methods.
Spec5.05c Hypothesis test: normal distribution for population mean5.05d Confidence intervals: using normal distribution

9 The leaves from oak trees growing in two different areas \(A\) and \(B\) are being measured. The lengths, in cm , of a random sample of 7 oak leaves from area \(A\) are $$6.2 , \quad 8.3 , \quad 7.8 , \quad 9.3 , \quad 10.2 , \quad 8.4 , \quad 7.2$$ Assuming that the distribution is normal, find a 95\% confidence interval for the mean length of oak leaves from area \(A\). The lengths, in cm, of a random sample of 5 oak leaves from area \(B\) are $$5.9 , \quad 7.4 , \quad 6.8 , \quad 8.2 , \quad 8.7$$ Making suitable assumptions, which should be stated, test, at the \(5 \%\) significance level, whether the mean length of oak leaves from area \(A\) is greater than the mean length of oak leaves from area \(B\). [9]

Question 9:
AnswerMarks Guidance
Answer/WorkingMarks Guidance
\(s_A^2 = \frac{(481.1 - 57.4^2/7)}{6} = \frac{521}{300}\) or \(1.737\) or \(1.318^2\)M1 A1 Estimate population variance using \(A\)'s sample (allow biased: \(1.489\) or \(1.22^2\))
\(57.4/7 \pm t\sqrt{(s_A^2/7)}\)M1 Find confidence interval
\(t_{6,\,0.975} = 2.447\ [or\ 2.45]\)A1 State/use correct tabular value of \(t\)
\(8.2 \pm 1.22\ [or\ 6.98,\ 9.42]\)A1 Evaluate C.I. correct to 3 s.f.
Population of \(B\) is Normal *and* has same variance as for \(A\)B1 State suitable assumptions
\(H_0: \mu_A = \mu_B,\quad H_1: \mu_A > \mu_B\)B1 State hypotheses
\(s_B^2 = \frac{(278.74 - 37^2/5)}{4} = 1.235\) or \(1.111^2\)B1 Estimate population variance using \(B\)'s sample (allow biased: \(0.988\) or \(0.994^2\))
\(s^2 = \frac{(6s_A^2 + 4s_B^2)}{10} = \frac{192}{125}\) or \(1.536\) or \(1.239^2\)M1 A1 Estimate population variance from combined sample
\(t = \frac{(57.4/7 - 37/5)}{s\sqrt{(1/7+1/5)}} = \frac{0.8}{0.726} = 1.10\ [2]\)M1 *A1 Calculate value of \(t\) (to 2 d.p.)
\(t_{10,\,0.95} = 1.812\ [or\ 1.81]\)*B1 State/use correct tabular value
\(\mu_A\) is not greater than \(\mu_B\)B1 Correct conclusion (AEF, dep *A1, *B1)
S.R.: Deduct only A1 if intermediate result to 3 s.f.
S.R.: Invalid method for \(t\) (max 6/9): \(t = 0.8/\sqrt{(s_A^2/7 + s_B^2/5)} = 0.8/0.704 = 1.14\)(M1)(A1)
Subtotal: 9Total: [14]
## Question 9:

| Answer/Working | Marks | Guidance |
|---|---|---|
| $s_A^2 = \frac{(481.1 - 57.4^2/7)}{6} = \frac{521}{300}$ or $1.737$ or $1.318^2$ | M1 A1 | Estimate population variance using $A$'s sample (allow biased: $1.489$ or $1.22^2$) |
| $57.4/7 \pm t\sqrt{(s_A^2/7)}$ | M1 | Find confidence interval |
| $t_{6,\,0.975} = 2.447\ [or\ 2.45]$ | A1 | State/use correct tabular value of $t$ |
| $8.2 \pm 1.22\ [or\ 6.98,\ 9.42]$ | A1 | Evaluate C.I. correct to 3 s.f. |
| Population of $B$ is Normal *and* has same variance as for $A$ | B1 | State suitable assumptions |
| $H_0: \mu_A = \mu_B,\quad H_1: \mu_A > \mu_B$ | B1 | State hypotheses |
| $s_B^2 = \frac{(278.74 - 37^2/5)}{4} = 1.235$ or $1.111^2$ | B1 | Estimate population variance using $B$'s sample (allow biased: $0.988$ or $0.994^2$) |
| $s^2 = \frac{(6s_A^2 + 4s_B^2)}{10} = \frac{192}{125}$ or $1.536$ or $1.239^2$ | M1 A1 | Estimate population variance from combined sample |
| $t = \frac{(57.4/7 - 37/5)}{s\sqrt{(1/7+1/5)}} = \frac{0.8}{0.726} = 1.10\ [2]$ | M1 *A1 | Calculate value of $t$ (to 2 d.p.) |
| $t_{10,\,0.95} = 1.812\ [or\ 1.81]$ | *B1 | State/use correct tabular value |
| $\mu_A$ is not greater than $\mu_B$ | B1 | Correct conclusion (AEF, dep *A1, *B1) |
| **S.R.:** Deduct only A1 if intermediate result to 3 s.f. | | |
| **S.R.:** Invalid method for $t$ (max 6/9): $t = 0.8/\sqrt{(s_A^2/7 + s_B^2/5)} = 0.8/0.704 = 1.14$ | (M1)(A1) | |

**Subtotal: 9 | Total: [14]**
9 The leaves from oak trees growing in two different areas $A$ and $B$ are being measured. The lengths, in cm , of a random sample of 7 oak leaves from area $A$ are

$$6.2 , \quad 8.3 , \quad 7.8 , \quad 9.3 , \quad 10.2 , \quad 8.4 , \quad 7.2$$

Assuming that the distribution is normal, find a 95\% confidence interval for the mean length of oak leaves from area $A$.

The lengths, in cm, of a random sample of 5 oak leaves from area $B$ are

$$5.9 , \quad 7.4 , \quad 6.8 , \quad 8.2 , \quad 8.7$$

Making suitable assumptions, which should be stated, test, at the $5 \%$ significance level, whether the mean length of oak leaves from area $A$ is greater than the mean length of oak leaves from area $B$. [9]

\hfill \mbox{\textit{CAIE FP2 2012 Q9 [14]}}