Warianty tytułu
Discovering Feed-backs in the Economy Using a Non-linear Correlation Coefficient
Języki publikacji
Abstrakty
W artykule podaje się propozycje stosowania do analizy zależności ekonopmicznych (z reguły nieliniowych, a nawet uwikłanych) asymetrycznej, nieliniowej miary korelacji, która spełnia swoją rolę także w przypadkach zależności uwikłanych. (fragment tekstu)
The process of economic growth can be described using either models as applied in mathematical economics or methods used by statisticians in their time series analyses. In the first case, the model is the expression of an a priori hypothesis on the structure of growth processes, whereas in the latter case use is made of the fact that statistical series provide a credible information on the type of stable and time-invariant elements of those developments. In the article, it is proposed that the economic interconnections be thought of in terms of relationships between pairs of variables but, instead of the classic correlation measure r/X,Y/ or r2/x,y/, using a variant of the latter that might be applied to the case of a non-linear stochastic relationship, to the article, such a measure has been defined and applied to two time series, one relative to the personal income and the other to the private consumption. The series examined appeared to exhibit sineoidal oscillations. Putting the latter together has led to an ellipse or spiral effect. This result means that a routine description of relationships between a pair of random variables using linear regression may lead to the belief that the growth in personal income is accompanied by a growth in consumption, whereas this must not be the case in the real world. The existence of oscillations in time series should be a serious warning that processes of economic growth do not develop correctly and the economy exhibits visible delaying effects. (original abstract)
Twórcy
autor
Bibliografia
Typ dokumentu
Bibliografia
Identyfikatory
Identyfikator YADDA
bwmeta1.element.ekon-element-000171277077