Contact us now



1. Expose the Meaning in Your Data

OPTIMAL STATISTICS(consulting) specializes in the design of research projects, experiments and surveys so as to maximize efficiency and the likelihood of significant results. Optimal design specifications can be counter intuitive and expert advice should always be sought before commencement even for relatively small projects.

Biological research and the social sciences are generating data at an enormous rate and it is imperative that the correct conclusions are reached. OPTIMAL STATISTICS does things the RIGHT way.

Examples (hypothetical):

  1. Researcher wishes to test the hypotheses that consumption of a certain food additive in childhood has an influence on later social functioning.
  2. Human resources manager wishes to find the most important workplace factors relating to productivity and harmony.
  3. Researcher wants to quantify the relationship between stress in childhood and later mental illness in a distinct sample.
  4. Researcher wants to determine the optimal drug dose in distinct populations.

2. Matching Data Using Probabilistic Algorithms

Organizations or researchers often face the difficult problem of either finding duplication within a data base or matching cases between different sources. Exact Matching falls far short of the optimal solution.

OPTIMAL STATISTICS provides matching services individually tailored to your databases to ensure the most accurate results. Databases vary and each requires an assessment of relative frequencies to ensure that the correct probability thresholds are used.

Some of the techniques used by OPTIMAL STATISTICS include:

  1. Metaphone classification
  2. Date classification
  3. Probabilistic assessment

The following simple example illustrates the concepts:

JOHN SMITH 20/08/73 and J. M. SMITH 20/08/73

Exact matching misses this although it is clearly a possible match for which a probability can be calculated.