It is only a contradiction if you ignore the measurement errors in BOTH. We already know the WHite dwarf methodology has an acknowledged error of 0.5 billion years. We don't know the error estimate for the observation in your linked article because it does mention it, but let's use 5% as a round number.
Taking the White dwarf estimate to it's upper most value (13.7 billion + 0.5 billion) yeilds 14.2 billion years. Taking your estimate of 15 billion years and using the minimum value based on a 5% measurement error, yeilds 14 and a quarter billion years.
Both measurements are within their error limits of each other.
In conclusion, there is no contradiction.