## A.5 Answer: TW 5 tutorial

1. In order from left to right: 70; 85; 100; 115; 130.
2. 30 points above the mean.
3. Two standard deviations above the mean.

A person with an IQ of 85 has an IQ that is 15 points below the mean.
This is equivalent to 1 standard deviation(s) below the mean IQ.
Using the 68--95--99.7 rule, this means that the proportion of the population with an IQ less than 85 is about 32%. In addition, the proportion of the population with an IQ above 85 is about 68%.

4. Answers vary. You probably cannot be very accurate using the 68--95--99.7 rule.
6. Two standard deviations from the mean is $$2\times 6.7 = 13.4$$, so 95% of females aged 18 and over have a measured height between $$161.4 \pm13.4$$ approximately, or from $$148.0$$cm to $$174.8$$cm.
7. As follows:
• $$z=(171-161.4)/6.7 = 1.43$$, so the probability is $$0.9236$$ or about 92%.
• So the odds are $$92.36/(100-92.36) = 12.1$$.
• The answer is just $$1-0.9236 = 0.0764$$ or about 7.6%.
• The probability of over 171cm is $$7.64$$%.
• So the odds are $$7.64/(100-7.64)= 0.084$$.
• $$z$$-scores are $$1.28$$ and $$2.78$$, so the answer is $$0.9973 - 0.8997 = 0.0976$$, or about 9.8%.
• So the odds are $$9.8/(100-9.8)=0.11$$.
8. Use the Tables: $$z=-0.84$$; then using unstandardising formula, the height is $$x=\mu+(z\times\sigma) = 161.4 + (-0.84\times6.7) = 155.772$$, or about 156cm.

1. Relative frequency.
2. Relative frequency.
3. Subjective.
4. Relative frequency.
5. Classical.

1. The decision-making process begin with making an assumption about the population parameter.
This means we know what to expect from the sample statistic. We never know exactly what value of the statistic we will see in the sample, because of sampling variation. But we can have some of idea of what values are reasonable to expect.
Then we take the sample (that is, we make the observations).
Then we compare the sample statistic that we observed... to the sample statistic we expected.
If what we observe is inconsistent with what was expected, then the the assumption is unlikely to be true.
However, if what we observe is consistent with what was expected, then the the assumption is probably true.

2. Step 1: Assumption about population parameter
Step 2: Expectation for sample statistic
Step 3: Observation of sample statistic
Decision: Consistent?
Conclusion A: Yes, supports assumption
Conclusion B: No, doesn't support assumption

1. $$z=(20-0)/10 = 2$$. The area or probability to the right is $$0.0228$$, or about $$2.3$$%.
2. The probability the SOI exceeds 20: $$2.3$$%. So odds: $$2.3\div(100-2.3)$$, or about $$0.024$$.
3. $$z=(-25-0)/10 = -2.5$$. The area to the left is $$0.0062$$, or about $$0.6$$%.
4. $$z=(-12-0)/10 = -1.2$$. The area to the right is $$0.8849$$, or about $$88.5$$%.
5. The two $$z$$-scores: $$(-10-0)/10 = -1$$ and $$(20-0)/10 = 2$$. The area between these is $$0.9772 - 0.1587 = 0.8185$$, or about $$81.9$$%.
6. The $$z$$ score corresponds to an area of $$0.80$$ to the left; from the tables, about $$z=0.84$$. This corresponds to an SOI of $$0 + (0.84\times 10)$$, or an SOI of about $$8.4$$.
7. The $$z$$-score is 0.385 from Table B.3, remembering that the area to the left would be 0.650 (draw a diagram!). So, the $$z$$-score is 0.385, so that the SOI values is $$x = \mu + (z\times\sigma) = 0 + (0.385\times 10) = 3.85$$.