I have an ion selective electrode but have a problem in calibration, it gives the results of the unknown samples around the final calibration concentration (point).
You can use a method of standard additives. But it is depend of possibilities of measuring device (pH-Ion meter or Potentiometer). many possibilities in use of Metrohm Titrino instruments. Metrohm propose many usefull applications for Na, K, Cl etc ions.
In addition to the references given by other authors, I suggest: Taking into account the actual sample matrix and interference ..
Depending on the design of selective electrode, perform tests with real samples and build their own calibration curve. Use the method of adding a solution of the analyte concentration and accurately if possible, use a certified standard solution.
Finally, experience with these tests will be useful to know well its selective electrode. The electrode manufacturer's suggestions are also very useful.
You can construct many calibration curves, depending of matrix, using an standard fabricated for some enterprise, example NIST.
Also, you can use the recommendations of own manufacturer. But, remember, you can manufacture the standards solutions on different matrices, following a very rigurous method. Including, with interferences analysis.
By giving values around the final calibration concentration, it sounds like a carryover problem. Are you sure you are carefully cleaning your electrodes and measurement cell after each measurement? Do you run your calibrants from high-to-low or low-to-high concentration? In the latter case, contamination from the calibrants could be the problem.