Tests for seasonality with a variable population at risk --------------------------------------------------------
^seast^ outcomevar popvar^,^ [ ^ed^wards ^ex^act ^not^ab ^gen(^newvar^)^ ^sec^ > tor^(^varname^)^ ^len^gth^(^varname^)^ ]
If ^edwards^ is used, the population variable need not be specified.
^sector()^ is required when the name of the sector variable is not ^month^.
Description -----------
^seast^ tests for seasonality of a binary outcome with a variable population at risk. Two tests are available, the Edwards test, and the Walter and Elwood test.
The default test is the Walter and Elwood test on a sector variable called ^month^. Otherwise the sector variable can be declared, as can its length. Months should be denoted by 1,2,...,12, where month=1 corresponds to January, etc. The default model takes no account of the differing lengths of month. An option is available where the number of days in a standard month (using the Gregorian calendar) is used.
Edwards test takes no account of the variation in the population at risk, and so does not require the population to be necessarily specified.
The Walter and Elwoods test takes account of the population at risk and the fact that it may vary between months.
Options for use with ^seast^ --------------------------
The default for ^seast^ is that there are 12 sectors, the sector variable is called ^month^, each month is the same length, and the Walter and Elwood test is to be used, which requires population data.
^sector(^varname^)^ declares the name of the variable denoting the sector. This should be numbered sequentially from 1,...,n, where n is the total number of sectors. If n=12, the program assumes that the sectors 1,..,12 correspond to January,...,December.
^length(^varname^)^ declares the name of the variable containing the length of each sector. For example if the sectors were the first 6 months of the year the length variable should contain the number of days in each of the months, i.e. 31,28,31,30,31,30. If ^length()^ is specified the test will take account of the length of the sectors varying.
^edwards^ specifies that the Edwards test is to be used. When using this option the population variable does not need to be included as the test does not utilise this information. The default Edwards test makes no adjustment for variable sector length.
^exact^ forces the test to adjust for variable month length. This is used when the sector has not been specified as the data consists of 12 months. This option sets each month length to be as used by the Gregorian calendar. ^exact^ does not need to be specified when the ^length()^ option is used.
^notab^ suppresses the displaying of a table of observed and expected number of events by sector.
^gen(^newvar^)^ generates a new variable containing the expected number of events for each month of birth. This may be useful for graphing the data, or for further calculations.
Examples of ^seast^ -----------------
. ^seast cases births^ . ^seast cases births, gen(exp) exact^ . ^seast cases births, sector(quarter) length(qlength)^ . ^seast cases, edwards^
Authors -------
Mark S. Pearce Department of Child Health,University of Newcastle upon Tyne. m.s.pearce@@ncl.ac.uk
Richard Feltbower Paediatric Epidemiology Group, University of Leeds.