The article is a lectio praecursoria held on May 18, 1968 at the University of Helsinki. It deals with some aspects connected with the adaptation of mathematic-statistical methods and in particular with multivariate methods. Among these regression, factor, and principal-component analysis are mostly used by the Finnish forest economists.
The PDF includes a summary in English.
This is a fourth paper in a series of studies concerning logging in farm forests. The objective was to construct a model representing the productivity per farm of logging for delivery cuts. The first objective was to find out how the productivity of logging should be measured. In the study, combined labour and capital are regarded as the input.
Second object was to consider what variables to use in theory to determine the productivity of logging for delivery cuts. The factors affecting productivity depend on the concept of productivity employed. The productivity per farm of logging in delivery cuts can be determined both by regional and by farm variables. Still considering solely the effect of the quality of labour and capital input, the variables representing the person in charge of the delivery cuts are important explanatory farm variables. Others represent the farm totality (size, lines of production etc.).
Third aim was to develop a statistical-mathematical method suitable for constructing the model. Possible methods include regression analysis, which is, however, not the best method when there is large number of different levels to explain, or factor analysis. The suitable method to use in constructing a model depicting the productivity of a farm, was considered to be to condense the explanatory variables into rotated orthogonal factors. After preliminary correlation analysis, estimates of the factors interpreted as rational were employed as the explanatory variables for selective regression analysis.
Last, the model was tried out against actual material collected per farm, and the hypotheses were tested.
The PDF includes a summary in English.