The TQI is a business criteria that averages all the technical Criteria. TQI is based on the grades of all modules on all technical criteria.
TQI is not based on technology level grade. Hence, we can not deduct its value from the technology level values
First, we compute the TQI of each module, which is done by a weighted average of the all technical criteria grades involved in the module (weighting is done according to the aggregate weight of the technical criteria for the TQI branch of the metric tree). If the computed grade exceed the calculated grade for a "critical" technical Criteria, then the grade takes the lowest value (In this case, the grade for a "critical" technical Criteria). Clearly, if we have an average of 3, but a critical technical Criteria has a value of 2, then TQI of the module will be 2 (This is important since you can improve some criteria leading to a higher average, if you have a critical technical criteria with a lower grade, the TQI value remains unchanged). Then, once all TQI grades for modules computed, we calculate the TQI of the application using a weighted average of TQI modules. The default weight is 1 for each module, but it is possible to define a metric weighting based on the technical size metric or on the business value. For more information about business Criteria computation and other computations, please refer to the official documentation. For 8.3, see this page: CAST AIP Measurement System explained
Queries related to TQI in the TKB are here:
The weighting of TQI is on a scale which is slightly different than other business criteria. The various technical criteria that make up TQI currently have default weights in the range of 1-22. This is due to trying to aggregate all technical criteria across various domains, as compared to areas which are in a similar domain which have weights that are between 1-9 (such as within the quality rule domain). This allows for a finer granularity of weights, and also allows for the ability to under-weight very minor aspects.
If you are reweighting any of the technical criteria that make up TQI for your purposes, then you may want to try to rank the technical criteria you are thinking of changing against the current ones to then see where you would assign the weight. For example, if you want to weight 'dead code' (current weight of 6) more than perhaps you would rank it in importance for your use as much as 'expensive call in loops' (current weight of 10), so that it would then have a new weight of 10 in your assessment model.