Methodology

Survey Design

The principle behind the survey design was a sampling frame based on standard industrial classification codes and number of employees in a company (SIC and size band). Each company within a survey cell was randomly selected, which helped to overcome the effects of any unforeseen bias. The Office of National Statistics was provided with the sample frame and was requested to provide the necessary optimal population, in this instance a total of 2210 companies. The results of the survey are estimated by the amount of waste produced per company in a particular cell in the sample frame. The results are grossed up to achieve a national figure from the representative population contained within each cell. As such each piece of sample data is used to estimate waste for the overall population of companies in Wales. The survey, therefore, included a measure of error, the sampling error, which is the difference between the true (the amount if 100% of companies in Wales had been surveyed) and estimated population value. The sampling frame, applying all of the above factors resulted in a relative standard error is 14.0%, giving a 90% confidence interval of +/- 23.0%. The sample represented a wide variety of commercial and industrial sectors but some were excluded, these are Agriculture, hunting and forestry, Construction and Demolition, Mining and quarrying and Fishing.

Data collection and reporting categories

13 surveyors conducted 2,144 on-site survey visits of companies across Wales with an employment range of 3 and above. A standardised questionnaire was used to collect data on the following:

Number of Employees;
Types of waste (EWC Code);
Quantity (either by weight or container type, size and quantity;
Whether waste was special;
The physical form of the waste;
The waste management option;
The waste contractor;
Initial and final destination of the waste; and
The cost of or income from the disposal.

The data was collected in predetermined categories to ensure consistency:

Standard Industrial Classification (SIC) was used to identify business sector - we grouped these into 22 groups for sampling (compared to the 32 groups used for the 1998/9 survey);

EWC (European Waste Catalogue) was used to categorise waste type – we used the Substance Oriented Classification of the EWC adopted for reporting under the EU Waste Statistics Regulation; this groups waste into 8 main categories and 13 subcategories;

A standard set of conversion factors to calculate weights where only data on volumes of waste were available;

Standard lists were also developed to define waste management options, container types, waste form (solid, liquid, sludge, etc.) and local authorities.

Data Cleansing and Validation

At the end of the survey, all data was transferred to the Environment Agency to undergo a further process of data cleansing. This was to ensure all waste streams were appropriate and fell within the definition of commercial and industrial controlled waste, that wastes amounts were not excessive for the type and size of industry. Volumes of waste were converted into tonnage using conversion factors based on EWC code. With tonnage figures for each waste stream a complex process of grossing up the figures to achieve total and national figures was conducted. The process involves calculating the total grossed up weight for each sector/size band, or the grossed up weight for each site, and then adding these together to calculate the grand total grossed up weight.

Response Rate

The final response rate was 97%. However, numerous replacements were required as a substantial proportion of companies in the sample had either ceased to trade or had moved to another location, perhaps either outside of Wales or the UK.

Estimation

A complex set of statistical calculations were developed to generate estimates of the types and quantities of waste produced by sector group and region/sub-region, methods of disposal and recovery used.

Accuracy of Data

This survey achieved a precision of +/- 3.5% at a 90 percent confidence level for total industrial and commercial waste. Users should therefore treat the information provided as the best estimate from a range