This question is from an amateur data analysis I did in May of data from Kepler.nasa.gov/Mission/discoveries. There were 80 Jupiter size planets in the Kepler "confirmed" planet table on this website (thru 4/1/15) that had host stars with metallicity data. Of these, 24 stars have [Fe/H}>0.2 and 40 have -0.1<[Fe/H]<+0.2. This is a much higher percentage than the sample as a whole. Would this be strong evidence that high metallicity measurements are due to accretion? Or, do the metallicity measurements given reflect abundance deeper into the star than accretion can count for (ie. higher metallicity stars are more likely to have Jupiters in general)?
This is a well known, well researched phenomena. Yes, there certainly is a correlation between metallicity and the likelihood of observing a hot Jupiter.
There are two classes of explanation.
(1) The correlation is real and due to the fact that it is easier to form planetary embryos from metal-rich material in the core-accretion model of giant planet formation.
(2) The observed exoplanets are those that weren't swallowed. The hosts are metal rich because they did swallow others.
Finally, it was thought possible that there could be an observational bias. Planets are easier to find and measure around high metallicity stars. This applied mainly to the Doppler technique, because high metallicity stars have stronger spectral lines. Transit detection should be much less affected.
Most think that some form of (1) is going on. (2) is argued against because the correlation still appears in stars with a wide variety of convection zone depths into which planetary material would have mixed. I'm not sure it is dead though, there have been a number of claims for abundance enhancements that look like the product of planet engulfment.
I'll add some references later. Nice work though to discover the correlation if you knew nothing of it previously.