VOA - Publications - CTR(E) IA - 261103 Mass Data Capture of Dwellinghouse Codes

VOA - Publications - CTR(E) IA - 261103 Mass Data Capture of Dwellinghouse Codes - Quality Control

CTR(E) IA - 261103 Mass Data Capture of Dwellinghouse Codes - Quality Control

These Instructions and Advice were prepared for internal use within the Valuation Office Agency as part of work towards a council tax revaluation in England that was due to take effect on 1 April 2007. The government announced its intention to postpone the revaluation on 20th September 2005 and preparatory work for a revaluation in 2007 has now been stopped.

This IA introduces amendments to IA CTR(E) 270303, paragraphs 2.8.3 – 2.8.7 – Data Extraction Quality Control and to paragraphs 2.10.1 – 2.10.8 – Data Input Quality Control.

1.	Introduction
1.1	Following discussion with Groups on the efficiency and speed of data capture and having taken advice from Inland Revenue Operational Research, in the interest of best practice across the network, Council Tax Revaluation Team have amended current instructions for quality control of extracted and inputted data.
1.2	With immediate effect the number of dwellings subject to the initial 100% quality check at the induction and training stage is reduced from 100 to 50.
1.3	It is stressed that the sample of 50 addresses should be randomly selected and be of varying types across a number of different streets. It is envisaged that both sets of quality control will be dealt with at the same time unless there is a significant time delay between the extraction and the input of the data.
2.	Action in Groups
2.1	Data Extraction CTR(E) 270303 paragraphs 2.8.3 – 2.8.7 should be replaced with the following revised instructions:
	2.8.3	Data extracted on the first 50 addresses by a staff member must be 100% quality controlled. Data extracted on to `Live CT Bandings - Dwelling Details Reports' needs to be checked against the source document. If a colour coding system has been used during the data extraction, Quality Controllers will know which source documents to quality control against.
	2.8.4	Any errors or omissions need to be recorded by the Quality Controller on the `Live CT Bandings - Dwelling Details Report'.
	2.8.5	If having quality controlled the data extracted on the first 50 addresses, more than 5% of addresses have data errors, or omissions where data was available, the Quality Controller will need to check 100% of the following set of 50 extracted by that staff member.
	2.8.6	100% quality control needs to continue until the staff member achieves a level where 5% or less of 50 addresses have errors. When this level is achieved a 2% sample of each batch allocated thereafter needs to be checked. The 2% of addresses to be checked should be randomly selected and if possible not properties in the same street.
	2.8.7	If at any time errors or omissions are found in the 2% sample then the staff member should revert to 100% quality control and only revert to 2% sampling when 5% or less of the addresses have errors.
	2.8.8	Addresses that have been selected for quality control should be marked on the `Live CT Bandings - Dwelling Details Report' lists by endorsing `EQC' (Extraction QC) against the checked address and be initialled by the Quality Controller with attention drawn to the data extractor of any errors and/or omissions.
	2.8.9	Errors and omissions should always be drawn to the attention of the staff member so that they are aware how to rectify these items in the future.
	2.8.10	A record of data extraction quality control must be maintained on the Data Extraction & Input Control Sheet of the data Capture Control File for the BA.
2.2	Data Input CTR(E) 270303 paragraphs 2.10.1 – 2.10.8 should be replaced with the following revised instructions:
	2.10.1	It is vital that extracted data is input as accurately as possible, therefore quality needs to be assured.
	2.10.2	It is essential that adequate training is provided to any staff member undertaking this task and that they have been familiarised with address selection and input screens. The data input on the first 50 addresses dealt with by a staff member must be 100% quality controlled. Data input on to the cdb needs to be checked against the `Live CT Bandings - Dwelling Details Report'.
	2.10.3	If having quality controlled the data input on the first 50 addresses, more than 5% of addresses have input errors, or omissions where data was available, the Quality Controller will need to check 100% of the following set of 50 input by that staff member.
	2.10.4	100% quality control needs to continue until the staff member achieves a level where 5% or less of 50 addresses have errors. When this level is achieved a 2% sample of each batch allocated thereafter needs to be checked. The 2% of addresses to be checked should be randomly selected.
	2.10.5	If at any time errors or omissions are found in the 2% sample then the staff member should revert to 100% quality control and only revert to 2% sampling when 5% or less of the addresses have errors.
	2.10.6	Quality controlled items should be marked on the `Live CT Bandings - Dwelling Details Report' by endorsing `IQC' (Input QC) against the checked address and initialled by the Quality Controller with attention drawn to the data inputter of any errors and/or omissions.
	2.10.7	Errors and omissions should always be drawn to the attention of the staff member so that they are aware how to rectify these items in the future.
	2.10.8	A record of data input quality control must be maintained on the Data Extraction & Input Control worksheet of the Data Capture Control File for the BA.

November 2003