**Allusion**-type message (of an artefact)- An ambiguous message which is not given with symbols with definite meaning and it therefore allows a certain range of interpretation. For example, allusion can be made by deviating from the normal shape or from the normal use of the product. Opposite: Statement-type message. See Product as Allusion.

**Analogy method**(of forecasting)- Method where the system to be predicted is assumed to develop similarly as another system, the development of which is known. See Analogy Method.

**ANOVA,**analysis of variance- Measuring the dissimilarity of two groups of observations by studying the variances of the groups. Analysis of variance

**Arithmetic scale**- A continuous scale of measurement, with evenly spaced steps. Examples of it are interval and ratio scales.Scales

**Artefact**or artifact (lat. artificium)- Anything made by man, a product of human work. Theory of Design

**Arteology**- Science which assists the design and production of artefacts. From Latin "ars" (art, technique) and Greek "logos" (word, knowledge). See Theory of Design.

**Artistic research**- Combining scientific and artistic methods in either investigation or in the creation of a work of art. Artistic research

**Artistic science**- Searching knowledge not only with the usual scientific methods, but also with suitable methods of arts. Artistic science

**Bias**- Systematic deviation between a sample and the population from which the sample was drawn. Opposite: (good) representativeness. See Sampling

**Chi test**- When the cases in a random sample are distributed into classes, chi test can be used to calculate the probability that a similar distribution exists in the population, too. Chi test

**Cochran Q test**- When two variables which have been measured from a random sample on a nominal scale have a certain relation between them, Cochran Q test can be used to calculate the probability that a similar relationship exists in the population, too. Cochran Q test

**Confidence interval**- The range of values of a population which includes with a certain probability the value measured in a random sample. Finding the confidence interval

**Conjoint analysis**- A questionnaire technique for finding the optimal combination of several attributes of the future product. Conjoint analysis

**Contingence**- A measure for a statistical association of two variables (usually of a nominal scale). See Contingence and Correlation on the page
*Quantitative Analysis*.

**Correlation**(product moment correlation or Pearson's correlation)- A measure for the linear association of two arithmetic variables. See Contingence and Correlation on the page
*Quantitative Analysis*.

**Dependent variable**- Series of measurements of the presumed effect of a cause. See Analysing Relationships between Variables.

**Descriptive study**- Disinterested study. Study which does not aim at improving the object nor later similar objects (as a contrast to normative study). See Planning an Empirical Study.

**Design theory**- Collection of permanent knowledge that is intended to assist the design of various new products. See Theory of Design.

**Diachronic study**- An inquiry of how the object of study has developed in time. Analyzing Development

**Diachronic work of art**- A work of art which must be enjoyed in a predefined succession; for example music, theatre or TV play presentation. See Beauty of Discovery. Opposite: a synchronic work of art.

**Discounting**- Finding the present value of a future payment by subtracting the interest due from today to the future date of payment. Benefits

**Disinterested study**- Descriptive study. Study which does not aim at improving the object nor later similar objects (as a contrast to normative study). See Planning an Empirical Study.

**Distinction**- Superioriry of an artist or his work, as compared to the normal level. The final assessment of a work of art as compared to expectations. Expectation and distinction

**Exemplar**- An earlier produced meritorious artefact which is used as a starting point in the design of a new product. See Theory of Design.

**Expectation**- What a member of the public of art expects to see or hear. Expectation and distinction

**Experience diary**- People report their actions on a special questionnaire. A variant of observation.

**Explanation**- Showing the reasons to a phenomenon or the state of the study object. Making it understandable. See Description and Explanation

**Exploratory study**- Studying an object about which very little is known and no theory is available. See Models in the Research Process.

**Extensive study**- Research project where the object of study consists of a class of cases. The target is to find invariances that exist in all or most of the cases. Contrary to intensive study. See Specific and General Knowledge.

**Extent**- The number of studied cases. See Planning an Empirical Study

**Gantt diagram**- "Road map", plan of future activities where several tasks are presented as stacked (usually horizontal) strings on calendar basis. Time Schedule

**Guttman scale**- A prefabricated series of statements about an attitude, arranged so that the first statements reveal a weaker conviction and the later ones a stronger attitude. Guttman scale

**Idiographic study**- Study of individual cases, as a contrast to nomothetic study which tries to find general laws). See Specific and General Knowledge

**Independent variable**- Series of measurements of the presumed cause of a phenomenon. See Analysing Relationships between Variables.

**Descriptive study**- Disinterested study. Study which does not aim at improving the object nor later similar objects (as a contrast to normative study). See Planning an Empirical Study.

**Intensive study**- Research project where the object of study consists of a single case or a few cases. Contrary to extensive study. See Specific and General Knowledge.

**Invariance**- A characteristic which is common to several or all cases in the material of study, in other words a pattern which does not vary. See Specific and General Knowledge, on the page
*Descriptive Theory*.

**Iteration**- A method of seeking the optimum by making successively very small changes to the object and measuring its utility after each change. Iteration

**Kitsch**- Immediately appealing art with no deeper content. Gradual comprehension

**Margin of error**- The values between which is the mean (or other interesting statistic) of the population, with the probability of (usually) 95%. The midpoint of the margin of error is the value measured in a random sample. Finding the confidence interval

**Mean, arithmetic**- The sum of all values divided by their number. Averages

**Median**- The value in the middle when all the values are arranged in a succession from the smallest to the largest one. Averages

**Methods engineering**- Study that aims at improving the productivity, output or safety of work. Methods engineering

**Mode**- The most common value among all the values. Averages

**Nomothetic study**- Study which tries to find general laws (as a contrast to idiographic study of individual cases). See Specific and General Knowledge

**Non-random sample**- A group which have been deliberately chosen from a population. Non-random sample

**Normative study**- Study which aims at improving the object or later similar objects (as a contrast to descriptive study, see Planning an Empirical Study.

**Participatory study**- Normative study where a few people from the interest groups personally assist in developing and evaluating the proposals of the project. An alternative to professional study. See Intensive normative approach.

**Population**- The class of cases, objects or events about which you want to gather data. See Demarcating the Study

**Professional study**- Normative study where the researcher gathers opinions from pertinent interest groups and uses them in developing and evaluating the proposals of the project. An alternative to participatory study. See Intensive normative approach.

**Random error**- Difference between one measurement and the mean of all the measurements. Measurement

**Range**- The difference between the smallest and the greatest value. Variation

**Reliability**- Repeatability. It shows how nearly the same are the repeated measurements of the same thing. Low reliability is the same as a large random variation or random error. Reliability

**Road map**- A Gantt diagram. Plan of future activities where several tasks are presented as stacked (usually horizontal) strings on calendar basis. Time Schedule

**Saturation**- When the study of new cases no more brings new knowledge to the researcher, the data is said to be saturated and no more material needs to be collected. Sample size

**Scientific art**- Art in which suitable scientific procedures are applied. Scientific development of a work of art

**Seasonal variation**- Among the variation types in a time series, seasonal variation is the one that follows some natural rhythm, e.g. the succession of years or days. Describing Development by Variables

**Self-reporting**- People report their actions on a special questionnaire. A variant of observation.

**Significance**(statistical)- It defines the probability that the measurements made from a random sample are true also in the population. Calculating statistical significance

**Standard deviation**- A method of describing the dispersion of measurements around their mean. The square root of variance. Dispersion of data

**Statement**-type message (of an artefact)- An explicit message, given with symbols with definite meaning that everybody understands. Example: a clockface. Opposite: Allusion-type message. See Product as Statement.

**Suggestion scheme**- An arrangement for judging and implementing the proposals made by employees. Suggestion scheme

**Synchronic study**- Study which is not interested in the development of the object. A cross-sectional study. Analyzing Development

**Synchronic work of art**- A work of art which stays invariable all the time. Its details can be studied in whatever succession the public prefers; for example a painting or an industrial product. See Beauty of Discovery. Opposite: a diachronic work of art.

**Systematic error**- The difference between the mean of all the measurements and the true value. A large systematic error means that the meter or the measurement method have a low validity. Measurement

**t test**- can be used to calculate the probability that a descriptive statistic, e.g. a mean or a correlation, that has been measured from a random sample, holds true in the population, too. t test

**Theory**- Description or explanation of empiria, usually written by researchers. Traditionally it is descriptive (see Descriptive Theory), but it can also be normative and help reaching practical targets, see Theory of practice.

**Theory of design**- Collection of permanent knowledge that is intended to assist the design of various new products. See Theory of Design.

**Theory of practice**- Knowledge that is needed in a profession. The skill of a metier in written form, collected through research. It can contain, for example, descriptions of good professional practice, relevant regulations and standards, instructions and algorithms. Planning an Empirical Study

**Time series**- A succession of values of a variable which have been measured with even intervals, e.g. daily or yearly. Describing Development by Variables

**Trend**- Long-time development; a temporary fashion or style, especially as compared to those of other times. See Other Descriptive Models of Development. In a quantitative study the trend often is defined as a linear growth or decrease of a variable which can be revealed by regression analysis, see Describing Development by Variables.

**Validity**- A successful definition of a theoretical concept so that exactly the right thing will be registered or measured. Absence of systematic error. Validity

**Variance**- The square of standard deviation. See Dispersion of data.

**Weak signal**- A factor that first seems insignificant when predicting a development, but gains importance during the time span of the forecast and ultimately becomes crucial. Applying a Causal Model

**Wilcoxon test**- When the variables which have been measured from a random sample on an ordinal scale have a certain relation between them, Wilcoxon test can be used to calculate the probability that a similar relationship exists in the population, too. Wilcoxon Test

- .

August 3, 2007.

Comments to the author:

Original location: http://www2.uiah.fi/projects/metodi