Analyse two-stage prevalence data

Analyse data from a two-stage survey to estimate prevalence. Analysis options are provided for probability-proportional-to-size (PPS) and simple random (SRS) sampling and for either stratified or unstratified design. Data for analysis must be submitted by copying and pasting from other software (for example Excel) in the specified format.

PPS sampling, Unstratified

For unstratified PPS sampling the following data is required:

• Two or more columns of data representing individual animal test results. The first column must contain cluster (village/farm) id for each animal and the second column contains individual test results (coded as TRUE or 1 for positive and FALSE or 0 for negative). A single row of column headings should be included with the data.

PPS sampling, Stratified

For stratified PPS sampling the following data is required:

• Two or more columns of data representing individual animal test results, as described above; and
• Two or more columns of cluster details. This data must include one record only per cluster represented in the testing data. The first column must be the cluster id corresponding to the animal testing data and a column with the heading "Stratum" must also be included, containing the stratun id for each cluster. A single row of column headings must be included with the data.

SRS sampling, Unstratified

For unstratified SRS sampling the following data is required:

• Two or more columns of data representing individual animal test results, as described above;
• Two or more columns of cluster details. This data must include one record only per cluster represented in the testing data. The first column must be the cluster id corresponding to the animal testing data and the second column must contain the total number of animals for each cluster. A single row of column headings must be included with the data; and
• The total number of clusters in the study area must also be entered in the appropriate input section.

SRS sampling, Stratified

For stratified SRS sampling the following data is required:

• Two or more columns of data representing individual animal test results, as described above;
• Two or more columns of cluster details. This data must include one record only per cluster represented in the testing data. The first two columns must contain the cluster id and the total number of animals for each cluster, as specified for unstratified SRS sampling. An additional column with the heading "Stratum" must also be included, containing the stratun id for each cluster, as for stratified PPS sampling. A single row of column headings must be included with the data; and
• Three or more columns of stratum details, with one record for each stratum in the data. First column is stratum id, as used in cluster data, second column is total number of clusters for the stratum and third is total number of animals for the stratum. A single row of column headings must be included with the data.

Download and check out the example data if you are unsure what is required.

Note:

• If one or more clusters have only single units sampled, within-cluster variance cannot be calculated and will be blank.
• within and between cluster variances differ slightly from the original Survey Toolbox version and are unverified.