August 26th

So with the data in excel (which you can learn how to export and where on the notes for August 24th), we spoke about what the average (using the midpoint) age of women would be at all age groups. We decided upon these because, for example, an age group of 16 would cover the girls age at the very day she turned 16, till the day before she turns 17, so an average age of 16.5 would be appropriate. Here is the image of all averages. We also named the group age to be able to quickly write a formula for the weights.

Long hand, what we are really doing is weighted averages (which is what demography is most concerned with).
Values:

(1)
$$Y_1,Y_2,...,Y_n$$

In Excel, we named the highlighted range, which contains the Y values that we want to average, as age

Weights:

(2)
$$W_1, W_2,...,W_n$$

(in Excel we did not label this, although we could have)

(3)
\begin{align} weighted average= \frac{W_1 Y_1 + W_2 Y_2 + ...W_n Y_n} { W_1 + W_2 + ... W_n} \end{align}

When you need to write in to get this on Excel you write

=sumproduct(weights, values)/sum(weights)

So, we have

=sumproduct(B7:B19, age)/sum(B7:B19)

which gives us our weighted average.

You do this for all values by copying the equation and pasting on rest of the slots in the column.

### Main Sources of Demographic Data:

Vital Registration or Statistical Registration Systems

• Birth Certificates
• Death Certificates
• Marriage Certificates
• Migration Registrations (used in some countries like Japan and Germany)
• Fetal Death Records
• Divorce
• Emigration

In theory, Statistical Registration Systems are:
Universal: all events are registered
Continuous: all times are covered

*Census is not continuous because it is not counting all the time.

*Sample or Survey may not be either because the until of recording is the event and not the name of the person (birth certificate is for the child, not the mother which then means the child is the event)

Next Lecture Aug 31

page revision: 14, last edited: 02 Sep 2010 12:27