## Info

(a) Draw an XY plot of the data and comment. (b) From the chart, would you expect the line of best fit to slope up or down In theory, which way should it slope (c) What would you expect the correlation coefficient to be, approximately (d) Calculate the correlation coefficient between margarine consumption and its price. (e) Is the coefficient significantly different from zero What is the implication of the result (The following totals will reduce the burden of calculation YY 67.52 YX 2101.70...

## The Poisson distribution

The section above showed how the Binomial distribution could be approximated by a Normal distribution under certain circumstances. The approximation does not work particularly well for very small values of P, when nP is less than 5. In these circumstances the Binomial may be approximated instead by the Poisson distribution, which is given by the formula where i is the mean of the distribution like 1 for the Normal distribution and nP for the Binomial . Like the Binomial, but unlike the Normal,...

## The x2 distribution

The x2 distribution with different degrees of freedom The x2 distribution with different degrees of freedom Estimating The sample variance is also a random variable like the mean it takes on differ a variance ent values from sample to sample. We can therefore ask the usual question given a sample variance, what can we infer about the true value To give an example, we use the data on spending by Labour boroughs in the example in Chapter 4 see page 151 . In that sample of 20 boroughs, the average...

## Is A 95 Confidence Interval Twice As Wide As A 47.5 Confidence Interval

Some of the more challenging problems are indicated by highlighting the problem number in colour. a Why is an interval estimate better than a point estimate b What factors determine the width of a confidence interval Is the 95 confidence interval a twice as wide, b more than twice as wide, c less than twice as wide, as the 47.5 interval Explain your reasoning. Explain the difference between an estimate and an estimator. Is it true that a good estimator always leads to a good estimate Explain...