Business Tendency and Consumer Survey Terminology
Ancillary activities are those that exist to support the main productive activities of an entity by providing non-durable goods or services entirely or primarily for the use of that entity. Ancillary activities are usually small scale compared to the principal activity they support. The output is always intended for immediate consumption within the same entity and therefore it is not usually recorded separately.
Balances (also called net balances) are used to summarise answers to multiple-choice questions in business tendency and consumer opinion surveys. "No-change" answers (such as "normal" or "same") are ignored and the balance is obtained by taking the difference between the weighted percentages of respondents giving favourable and unfavourable answers.
There are several ways to measure "the business cycle". In general, it is defined as fluctuations in the level of economic activity around an estimate of its underlying trend. Starting from a position where the level of activity is close to the trend level, a business cycle is said to be complete when the level of activity has returned back to the trend following a period when it was above, then below, the trend level, or vice-versa.
Composite leading indicator.
A leading indicator is an economic time-series series which exhibits a leading relationship with respect to the turning points of a reference series. A composite leading indicator is constructed by combining two or more leading indicators into a single index number.
Given a pair of related variables (X and Y), the correlation coefficient (r) measures the degree to which they co-vary in a linear fashion. r will be positive when the two variables move in the same direction and negative when the two variables move in opposite directions. r takes values from -1 to +1. Ignoring the sign, the larger the value of r the greater the strength of the association between the two variables.
Correlation is the tendency for two variables to both move in the same or opposite directions.
Cross-correlation of business cycles
A measure of how closely aligned the timing of movements in activity are for two countries over their business cycles. The cross-correlation statistic can range from -1 to +1. In general, the closer the cross-correlation is to the value of +1 the more in phase and synchronised the business cycles will be. A value of -1 would indicate that the two series move perfectly in a counter-cyclical direction. A value near zero indicates that there is no statistical relationship between the series.
A cycle in economic activity is the time span separating two turning points of the same nature (two peaks or two troughs).
A diffusion index provides a summary of answers to multiple-choice questions in business tendency surveys. It is defined as the fraction of favourable answers plus half of the fraction of no change answers. Diffusion indices are alternatives to "Balances" as a way of summarising answers to multiple-choice questions.
An establishment is an enterprise, or part of an enterprise, that is situated in a single location and in which only a single (non- ancillary) productive activity is carried out or in which the principal productive activity accounts for most of the value added.
As used in business cycle analysis, it is a measure of the extent to which the current level of economic activity depends upon the level in the past. The higher the degree of dependence the longer or more persistent the business cycle will be.
When a survey is repeated on several occasions, but observations are taken on the same sample instead of a new sample for each occasion, the sample is said to be fixed.
A second interview, - e.g. by telephone or personal visit - to obtain missing data or clarify original responses.
Growth cycles are recurrent fluctuations in the series of deviations from trend. Growth cycle contractions include slowdowns as well as absolute declines in activity, whereas business cycle contractions include only absolute declines (recessions). The OECD Cyclical Indicator System tries to predict growth cycles rather than business cycles.
A kind-of-activity unit is an enterprise, or a part of an enterprise, in which the principal productive activity accounts for most of the value added.
The correlation observed between two series where one of the series has been lagged with reference to the other.
An event occurring at time t+k (k>0) is said to lag behind event occurring at time t, the extent of the lag being k. An event occurring before another is said to have a negative lag.
An economic indicator whose peaks and troughs during the business cycle tend to occur sooner than those of the general economy.
A local unit is an enterprise, or part of an enterprise, which engages in productive activity at one location. A local unit may have more than one kind of economic activity.
MCD - Months for Cyclical Dominance
The MCD is defined as the shortest span of months for which the I/C ratio is less than unity. I is the average month-to-month changes (without regard to sign) of the irregular component of the series and C is the trend-cycle component of the series. For quarterly series, there is an analogous measure, Quarters for Cyclical Dominance (QCD). MCD and QCD are used to determine the minimum number of months or quarters that need to be included in a moving average to eliminate irregular fluctuations.
When unqualified, the mean is usually interpreted as the arithmetic average. Other means are the geometric and harmonic.
Non-response errors occur when the survey fails to get a response to one, or possibly all, of the questions. Non-response causes both an increase in variance because of the decrease in the effective sample size and because some method must be used to impute a response. Imputation methods may cause bias if the non-respondents and respondents differ with respect to the characteristic of interest.
In sample surveys, the failure to obtain information from a designated individual or unit for any reason (death, absence, refusal to reply) is termed non-response. The proportion of non-responses in the total number of individuals or units included in the sample is called the non-response rate.
An error in sample estimates which cannot be attributed to sampling fluctuations (see Sampling Error). Such errors may arise from many different sources such as non-response, defects in the frame, faulty demarcation of sample units, defects in the selection of sample units, mistakes in the collection of data due to misunderstandings, to negligence or dishonesty on the part of the investigator or of the interviewee, mistakes in editing or processing the data, etc.
Observation errors are components of non-sampling error and are generated during the data collection process. They may occur because the question is wrongly worded or misleading, because the interviewer makes an error whilst asking the question or because the interviewer records incorrectly the answer given by the respondent.
Over-coverage means that some units that are outside the target universe have been included in the survey frame.
Business tendency surveys usually cover a group of enterprises that is kept fixed for two or more rounds of the survey. The group is referred to as a panel.
An observation in an ordered series is said to be a "peak" if its value is greater than the value of its two neighbouring observations.
Time span between a peak and a trough.
As defined in the SNA, the principal activity of a producer unit is the activity whose value added exceeds that of any other activity carried out within the same unit (the output of the principal activity consists of goods or services that are capable of being delivered to other units even though they may be used for own consumption or own capital formation).
See Random Sampling
Processing errors are components of non-sampling error and are generated during data processing.
The data collected in business tendency and consumer opinion surveys are termed "qualitative" because respondents are required to assign qualities to the items of interest. For example, they may be asked to say whether their order books are "higher", "lower" or "same" compared with some previous period. The qualitative data obtained in business tendency and consumer opinion surveys are also described as "categorical" because respondents are required to choose between two or more response categories, such as "better", "same", "worse".
The data collected in most statistical surveys are quantitative in contrast to the qualitative data collected in business tendency and consumer opinion surveys. Quantitative data are expressed in numbers, tons, litres, amounts of expenditures, etc.
A method of sampling in which the sampling units are selected according to a number of controls designed to ensure that the units sampled are broadly representative of the target population. In the case of a consumer opinion survey the controls might refer to age, sex, income group and education level. In the case of a business tendency survey the controls might refer to size, location, and kind of activity.
Any method of selecting a sample based on the theory of probability. At any stage of the process of selection the probability of any set of units being selected must be known. It is the only general method known which can provide a measure of precision of the estimate. Sometimes the term random sampling is used in the sense of probability sampling.
Cyclical indicator systems are constructed around a reference series. For a system of leading indicators, the reference series is the economic variable whose cyclical movements it is intended to predict. Indices of industrial production and gross domestic product are often used as reference series.
These are the units or individuals about which information is sought. In the case of business tendency surveys, reporting units are usually establishments or kind of activity units.
These are the persons who complete the questionnaires that provide the information about the reporting units. For business tendency surveys the response units are, ideally, the managers of the establishments or kind of activity units that constitute the reporting units.
The number of sampling units which are to be included in the sample. In the case of a multi-stage sample this number refers to the number of units at the final stage in the sampling.
That part of the difference between a population value and an estimate of that value which has been derived from a random sample, that is due to the fact that only a sample of values is observed. The totality of sampling errors in all possible samples of the same size generates the sampling distribution of the statistic which is being used to estimate the parent value.
A sampling frame is a body of information about the population being investigated which is used as the basis for selecting samples and in subsequent estimation procedures. For business tendency surveys the sampling frame may be a business register maintained by a national statistical office or the membership list of a trade organisation.
Sampling units are the units - typically persons or enterprises - described in the list or register from which the sample is selected. For business tendency surveys, the sampling units are usually enterprises listed in a business register. This term is synonymous with "Survey Units".
Seasonal adjustments are made to economic time series to eliminate the effects of seasonal factors that affect economic activity. Examples of such factors are public holidays and the seasonal farming cycle.
Simple random sampling
Sampling in which every member of the population has an equal chance of being chosen.
Stratification is the division of a population into parts, known as strata. The strata should be selected in such a way that the members of each stratum are relatively homogeneous with respect to the variables that are to be measured. Stratification may be done by kind of activity or size in the case of enterprises or by social class or income in the case of households.
A sample selected from a population which has been stratified, part of the sample coming from each stratum. For business tendency surveys, the strata are usually defined by kind of activity and by size of enterprise.
See Sampling unit
The target population is the entire group of individuals, households or enterprises whose characteristics are to be measured by means of a sample survey.
In time series analysis, a given time series can be decomposed into a trend component, a cyclical component, a seasonal component and an irregular component. An essential feature of the concept of trend is that it is smooth over periods that are long in relation to the unit of time for which the series is recorded. A trend is usually represented by some smooth mathematical function such as polynomials in the time variable or by moving averages.
A turning point is an observation in a time series which is a peak or a trough. When several contiguous values are equal, and greater than or less than the neighbouring values, a convention is required in order to determine which is regarded as the turning point, e.g. the middle one may be chosen.
A statistic is said to be an unbiased estimate of a given parameter when the mean of the sampling distribution of that statistic can be shown to be equal to the parameter being estimated.
A sample drawn by a method which is free from bias. This implies not only freedom from bias in the method of selection, e.g. random sampling, but freedom from any bias of procedure, e.g. wrong definition, non-response, design of questions, interviewer bias, etc.
Under-coverage means that some units that should be included in the survey frame have been omitted.