Index  A combination of weighted individual indicators, used to measure change without giving an actual numerical value. An index is usually given as a percentage, with the base period set to 100. 
 
Information development plan (IDP)  A systematic means of cataloguing and addressing priority needs for data in a particular field. 
 
Input data item  A particular characteristic of units in a population which is measured or observed and which is obtained from the respondent. 
 
Input editing  The checking of individual responses to a survey prior to aggregation (before, during or after data entry). 
 
Integration  The structuring of data to enable it to be used beyond the immediate purpose for which it was produced. 
 
Interpretability  A quality measure of the degree to which statistical information can be understood, explained and used. 
 
Linear regression  A mathematical equation showing how one independent and one dependant variable for which the relationship between the variables is approximated by a straight line. 
 
Logical edit  A check to ensure that two or more categorical items in a record do not have contradictory values, for example a respondent being 16 and receiving the age pension. 
 
Longitudinal dataset  A set of statistical data which observes the same analysis units over a substantial period of time. 
 
Back to Top of Page  
Mean  The arithmetic average of a set of values. 
 
Median  The middle value of a set of values that have been sorted in order. 
 
Metadata  Information about statistical data, for example contact person, how to access, accuracy, time period covered, scope, definitions. 
 
Methodology  The statistical theory (logic, principles and procedures) used to devise and conduct a statistical collection. See Collection methodology, Sampling methodology. 
 
Microdata  Statistical data of the smallest level. See unit record. 
 
Missing data edit  A check that data that should have been reported were in fact reported. 
 
Model  An abstract mathematical problem that approximately corresponds to the real world problem. 
 
Modelling  The process of drawing together several variables and data sources to make inferences about the relationships between the variables. 
 
Nonrespondent  An individual, or representative of an organisation, who does not provide information when requested in a survey. 
 
Nonsampling error  Errors that occur in producing statistical information that are not caused by sampling methodology. For example, errors can occur from the respondent, questionnaire, interviewer, processing, editing, field procedures, frame undercoverage, et cetera. 
 
Optical character recognition (OCR)  A form of imaging where the hand written responses on a survey form are extracted and converted to a useable form. The process involves capturing the electronic image of the form, then converting in into a useable format via repair and interpretation processes. 
 
Optical mark recognition (OMR)  A form of imaging where the response marks on a survey form are read and interpreted to give useable data. 
 
Output data item  A particular characteristic of units in a population which has been collected and processed and is intended for dissemination. It may be an aggregate of data items. 
 
Output editing  The checks applied to aggregate data once sampling weights have been applied. 
 
Outputs  Reports, graphs, tables, publications, et cetera. displaying the aggregate data from one or more collections. 
 
Parameter  The value of a characteristic of a complete population. 
 
Percentile  A numerical measure that also locates values of interest in a dataset. 
 
Pilot test  A pilot test is a small scale quantitative test of a survey, used to evaluate one or more of its components, such as the questionnaire or processing. 
 
Population  All the individuals or groups about which information is required, that is, the complete set of objects of interest in a statistical collection. A population may share a common set of characteristics. 
 
Postenumeration  The time in the process of a statistical collection after the data have been collected. 
 
Postenumeration survey  A study of a sample of respondents and nonrespondents after information has been collected, with the aim of evaluating the quality of the data. This can occur after a pilot test or after the final survey. It questions how the respondent completed the form and gauges comprehension of survey concepts. 
 
Processing  The systematic transformation of collected data into statistical information such as tables and graphs. 
 
Back to Top of Page  
Qualitative  Of interpretive or descriptive information based on opinions, perceptions, feelings and beliefs. 
 
Quality  The standard of an output, assessed in terms of relevance, accuracy, timeliness, accessibility, interpretability and coherence and described in information (metadata) accompanying the output. 
 
Quality assessment  A judgement on the quality of a data source with respect to a specific data need. It may be derived from a Quality Declaration. 
 
Quality declaration  A statement about a data source, addressing dimensions of the data quality framework. 
 
Quality framework  A template of the measures of quality, which is used to assess overall fitness for purpose. 
 
Quantitative  Of information based on numerical data 
 
Range  The difference between the largest and the smallest value. 
 
Raw data  Data as provided by the respondent, before editing, weighting and aggregation. 
 
Regression  A technique for predicting the value of a variable from values of one or more other variables. 
 
Relationship  The way in which two or more characteristics are connected. 
 
Relative Standard Error  A measure of accuracy of a survey estimate, formed by the ratio of standard error to the estimate and often expressed as a percentage. 
 
Relevance  A quality measure relating to the pertinence of statistical output to the original objectives or the user's objectives. 
 
Reliability  The extent to which a survey would produce the same results, if repeated using another statistically equivalent sample and methodology. 
 
Respondent  An individual, or representative of an organisation, who provides information when requested in a survey. 
 
Respondent load  The effort, in terms of time and cost, required for respondents to provide satisfactory answers to a statistical collection. 
 
Response rate  The percentage of the target sample or population from which responses or usable data were obtained. The response rate can apply to the whole survey form or to the individual questions. 
Back to Top of Page  
 
Sample  The part of a population which is selected in a survey, for the purpose of studying characteristics of the entire population of interest. 
 
Sample design  A set of specifications which describes the population, frame, survey units, sample size and sample selection method for a particular survey. 
 
Sampling error  The error which arises because the data are collected from a part, rather than the whole, of the population. 
 
Sampling methodology  The principles and procedures involved in selecting a representative subset of a population for the purpose of drawing conclusions about the population. See Methodology. 
 
Scope  The definition of the population of units about which information is required, for example households with children, retail businesses, people over the age of 65. 
 
Seasonal adjustment  The removal of the estimated effects of normal seasonal variation from a time series so that the effects of other influences can be more clearly recognised. 
 
Selfenumeration  A method of data collection in which respondents complete the survey questionnaires without the involvement of an interviewer, for example in a mailout survey. 
 
Sensitive cell  A cell of a table containing a value which could lead to the identification of a respondent. 
 
Sensitivity  (1) The degree to which data could change as a response to changes in assumptions or methods used in compilation. (2) The propensity of respondents to react to questions on a personal topic in a way that may adversely affect data quality. 
 
Sensitivity analysis  (1) A method of analysing the sensitivity of a decision to a change in one or more of the assumptions used in making it. (2) An investigation of the degree to which the statistical data is affected by a change in the value of some parameter or variable, or by a combination of changes. 
 
Simple random sampling  A method of sampling where each member of a sampling frame has an equal chance of selection and each possible sample of a given size has an equal chance of being selected. 
 
Stakeholder  Any individual or group with an interest in a particular collection. The interest may relate to the conduct of the collection or to its objectives and outcomes. 
 
Standard  An accepted rule or definition for data items, collections, classifications, et cetera. 
 
Standard deviation  The positive square root of the average of the squared differences of a set of measurements from their mean. 
 
Standard error  A measure of the variation among the estimates from all possible samples, and thus a measure of the precision with which an estimate from a particular sample approximates the average results of all possible samples. The unit of measurement for the standard error is the same as the variable of interest. 
 
Statistic  The summary value of a variable or attribute calculated from sample data. 
 
Statistical activity  (1) The designing or running of a census, survey or administrative data system. (2) Extracting or using data from a census, survey or administrative data system. 
 
Statistical Clearing House (SCH)  An independent body which is the central clearance point for all Australian Commonwealth Government surveys involving 50 or more businesses. 
 
Statistical collection  Selfcontained statistical activity involving the gathering, processing and combining of data on a particular theme. 
 
Statistical cycle  The sequential process of gathering, analysing and disseminating information on a particular theme. 
 
Subject matter  The theme or area of interest for a particular survey or collection. 
 
Summary measures  Measures of location (for example mean), spread (for example range), aggregation, et cetera of a set of data. 
 
Survey  The collection of information about characteristics of interest from some, or all, units of a population using welldefined concepts, methods and procedures. 
Back to Top of Page  
.  
Tabulation  A counting of the number of cases falling into each of a number of categories. 
 
Time series  A statistical record of a particular activity where the data is measured at regular intervals over a period of time, for example monthly unemployment rate. Time series are collected on this basis to assist understanding of the current situation, enabling the most recent data observations to be placed in a meaningful historical perspective. 
 
Timeliness  A quality measure relating to (1) the time taken between the occurrence of the characteristics/events being measured and the release of statistical output and (2) whether the output of a collection is sufficiently uptodate for the user's purpose. 
 
Unit  The entity used in the design, collection, compilation, tabulation or publication of statistical data. There are many types of unit. See Analysis unit, Reporting unit, Sampling unit. 
 
Unit record file  A list of data relating to individual population members. 
 
Validation  The checking of data values to ensure consistency either within the data source or with data from different sources. 
 
Validation edit  A check of the validity of a basic identification or classificatory item in unit data, for example sex is coded only as one of M or F. 
 
Variable  A characteristic that may assume more than one of a set of values to which a numerical measure can be assigned (for example income, age, weight, et cetera.). 
 
Variance  The arithmetic mean of the squared deviations from the population mean. 
 
Weight  The number of units in a population represented by a particular unit in a sample (for example a weight of 20 means that the sampled unit represents 20 units in the population). 
 
Weighting  The procedure for applying weights to survey results
