Microsoft Office Tutorials and References

In Depth Information

Although there are many questions that can be asked, these four are important

and will allow us to investigate numerous analytical tools in
Data Analysis
.Asa

note of caution, let us keep in mind that this example is based on a very small

amount of data; thus, we must be careful to not overextend our perceived insight.

The greater the amount of data, the more secure one can be in his observations. Let

us begin by addressing the ﬁrst question.

3.5.1 Graphical Analysis

Our graphical analysis of the sales data has already revealed the possibility of
sys-

tematic behavior
in the series; that is, there is an underlying
system
that inﬂuences

the behavior of the data. As we noted earlier, all series, except for product D, display

some form of cyclical behavior. How might we determine if systematic behavior

exists? Let us select product E for further analysis, although we could have chosen

any of the products.

In Exhibit 3.18 we see that the product time series does in fact display repeti-

tive behavior; in fact, it is
quite
evident. Since we are interested in the behavior of

both the yearly demand and quarterly demand, we need to rearrange our time series

data to permit a different type of graphical analysis. Table 3.3 shows the data from

Table 3.1 in a modiﬁed format: each row represents a year (1–6) and each column

a quarter (1–4); thus, the value 101 represents quarter 3 in year 5. Additionally,

the right most vertical column of the table represents yearly totals. This new data

conﬁguration will allow us to perform some interesting graphical analysis.

Exhibit 3.18
Product E time series data

Search JabSto ::

Custom Search