As the quotes in this release are based upon info associating with a sample of companies rather than a full enumeration, they go through sampling variability. That is, they may vary from the price quotes that would have been produced if the info had actually been gotten from all employers. This difference, called sampling mistake, must not be puzzled with mistake that may occur because of imperfections in reporting by participants or in processing by the ABS. Such mistake is referred to as non-sampling mistake and may take place in any enumeration whether it be a full count or sample. Efforts have been made to reduce non-sampling mistake by careful style of surveys, comprehensive monitoring of returns and quality control of processing.

The tasting error connected with any price quote can be approximated from the sample results. One procedure of sampling error is offered by the basic mistake which shows the degree to which an estimate might differ from the worth which would have been obtained from a full enumeration (the 'true worth'). There have to do with two possibilities in three that a sample estimate varies from the true worth by less than one standard mistake, and about 19 chances in 20 that the distinction will be less than two basic errors.

An example of using a standard mistake on levels is as follows. If the estimated variety of job vacancies was 25,000 with a standard mistake of 2,500, then there would be about two possibilities in three that a complete enumeration would have provided a quote in the variety 22,500 to 27,500 and about 19 possibilities in 20 that it would be in the range 20,000 to 30,000.

An example of using a basic error for a quarterly modification price quote is as follows. If the projected basic error for a quarterly modification estimate of job vacancies was 1,000 and the quarterly change price quote between two quarters was 4,500, then there would be about 2 opportunities in 3 that a complete enumeration would have provided a quarterly modification estimate in the range +3,500 to +5,500 and about 19 possibilities in 20 that it would be in the variety +2,500 to +6,500.

Quarterly motions in price quotes of job vacancies are considered to be statistically significant where they surpass two standard errors.

Another procedure of the sampling error (for level approximates just) is the relative basic error, which is gotten by revealing the basic error as a portion of the quote to which it refers. Level approximates with a relative basic mistake in between 25% and 50%, denoted by an asterisk in this release, are subject to tasting variability normally considered to be expensive for most useful functions and should be used with caution. Level estimates with a relative standard mistake of 50% or more, denoted by a double asterisk, are considered to be too undependable for general use.

The following two tables shows the standard errors for quarterly level and motion for States and areas by Sector, based upon original data for the current quarter. The third table shows the basic errors for level estimates by industry.