About the algorithm

The underlying algorithm for calculation of job effects on the website Jobeffekter.dk.


When you search for a job effect for an intervention and target group on Jobeffekter.dk, it happens by means of an algorithm that the Agency for Labour Market and Recruitment has developed in collaboration with researchers. 

The algorithm starts by finding several study results, that partly or totally matches your chosen search criteria. Be aware that there can be more study results from each unique study. Every study result gives a knowledge contribution based in the study results quality and relevance calculated through a combination factor and researchers' weighting of the studies.

The knowledge contribution of the study results are aggregated to determine the complete stock of knowledge. The stock of knowledge can be between 0 and 1, where 1 is obtained by the highest relevance and quality. In this instance, a score of 1 could be from a Danish study of high quality, based on a Randomized Controlled Trial and published in a top-tier scientific journal.

The following gives an example of how the knowledge contribution of a study result is calculated.

Example

The researcher has uploaded a study estimating a positive effect of a business-oriented intervention aimed at insured unemployed individuals. The study is an effect analysis using register-based data for Denmark. The results have not been published in a scientific journal, and the researcher has assessed the quality of the study as satisfactory.

In this case, the weighting of the study result will be as follows, if the search is limited to business-oriented interventions aimed at insured unemployed individuals:

Criterion

Characteristics

Weight

Quality assessment

Satisfactory

0.5

Analytical method

Econometric/statistical effect study

0.8

Geography

Denmark

1

Published in a scientific journal

No

0.5

Combination factor

No

1

Weighting of multiple results

The study yields two results in the category “Partial target group results”

0.5

Knowledge contribution of the result*

0.1

*Total weighting

Consequently, this result will be included with a weight of 0.10 for a positive effect in the calculation of the job effect of a business-oriented intervention aimed at insured unemployed individuals.

Please note that the weighting of geography differs, when using the Danish or English version of Jobeffekter.dk. In the Danish version a study from Denmark is weighted 1, a study from North Europe is weighted 0.75 and a study from a OECD country is weighted 0.5. In the English version all studies from Europe (including Denmark) or a OECD country is weighted 1.