Microsoft Office Tutorials and References

In Depth Information

**What’s in the Future: Using FORECAST, TREND, and GROWTH to Make Predictions**

What’s in the Future: Using FORECAST,

TREND, and GROWTH

to Make Predictions

The FORECAST function does just what its name suggests — it forecasts

an unknown data value based on existing, known data values. The function

is based on a single important assumption — that the data is linear. What

exactly does this mean?

FORECAST

The data that FORECAST works with are in pairs — there’s an X value and a

corresponding Y value in each pair. For example, perhaps you’re

investigating the relationship between people’s heights and their weight. Each data

pair would be one person’s height, the X value, and their weight, the Y value.

Many kinds of data are in this form — sales by month, for example, or income

as a function of educational level.

You can use the CORREL function to determine the degree of linear

relationship between two sets of data. See Chapter 9 to find out about the CORREL

function.

To use the FORECAST function, you must have a set of X-Y data pairs. You

then provide a new X value, and the function returns the Y value that would

be associated with that X value based on the known data. The function takes

three arguments:

✓
The first argument is the X value that you want a forecast for.

✓
The second argument is a range containing the known Y values.

✓
The third argument is a range containing the known X values.

Note that the X and Y ranges must have the same number of values or the

function returns an error. The X and Y values in these ranges are assumed to

be paired in order.

Don’t use FORECAST with data that isn’t linear. Doing so produces inaccurate

results.