BS ISO 8559-3:2018 pdf download – Size designation of clothes Part 3: Methodology for the creation of body measurement tables and intervals
4 Principles related to the methodology of the creation of the body measurement tables
4.1 General In order to get a garment fitting correctly on a body, it is essential to collect the body dimension measurements.
These measurements are useful to design the 2‑dimensional garment pattern, which are to be assembled in garment production. These measurements are useful also to design the 3‑dimensional garment, such as seamless knitted garments. Within the clothing industry, when faced with the need to mass produce clothing articles intended to fit a population of varying morphologies, a series of body measurement tables are internally determined in order to design the appropriate clothing articles. The creation of body measurement tables is based on a statistical approach of the body dimensions of the population.
The producer/retailer strategy, taking into account its market location, its targeted population, etc. leads to the creation of its own body measurement tables.
Nonetheless, whatever the basis of the body measurement tables, their creation is based on the same principles as described in this document. Any regression formula calculated for one group of population is specifically applicable to this group and should not automatically be applied to other groups. Annex A gives an overview of the clothing industry needs regarding the anthropometric data.
Annex B shows an example of a body measurement table content.
Annex D gives an overview of key statistical tasks.
4.2 Selection of the explanatory variables and the dependent variable (statistical methodology) The explanatory variable refers to the selected variable input data in the measurement tables. They are mainly based on a dimension characterizing a body measurement (as defined in ISO 8559‑1) or based on other body characteristics (e.g. body mass). Annex C describes statistical models commonly used for the creation of body measurement tables. The selection of the explanatory variable is based on various criteria:
— the two or more selected explanatory variables, representing body dimensions, shall be statistically independent and perpendicular: one representing the measurement on the vertical axis and the others that of width or girth (on the horizontal axis);
— a robust correlation does exist between the explanatory and dependent variables;
— based on the given ranges and/or intervals related to the development of clothing, the number of sizes is led by the combination of the chosen dimensions
— the values of the measurements are easy to remember by the consumer. The selection of the explanatory variables shall be carried out through various statistical studies, with the successive addition of dimensions to the statistical model.
This allows the best combination of the dimensions for the explanatory variables and the dependent variable to be obtained. It is important to take into consideration the following questions while selecting the explanatory and dependent variables in order to design and develop clothing that fits as many target consumers as possible.
a) What are the standard deviations of the dimensions selected as the dependent variable in relation with the dimension(s) selected as the explanatory variables
b) What should be the relation between the standard deviations mentioned in a) and the tolerance of clothing fitting
The selection of explanatory variable is usually from the few most important dimensions, such as height, chest girth, waist girth, hip girth. And then multiple regressions lead to calculate a residual standard deviation, which provides information not explained by the combination of explanatory variable on the prediction of dependent variable. Lower is the residual standard deviation; more satisfactory is the combination of the main dimensions regarding the prediction of the related dependent variable. The residual standard deviation is comparable with the clothing fitting tolerances. The point is that the residual standard deviation leads to calculate a difference allowing for a measured value data, i.e. 95 % of concerned people at this value are placed in this difference, to be compared with the tolerances.