Hey, just to reiterate what has been said before, the results of the flame test and the precipitate test are the ones to consider here (what element gives the yellow flame colour (possibly "golden yellow") and which likely anion gives a positive result in the ppt test with silver nitrate?!) As for your collegues seeing flecks of purple in the flame...they may well have! I assume this test was performed using a bunsen burner. Bunsens can often give erronious flame colours since old powders from other similar tests can fall into the burner. It could be that the flecks of purple observed were the result of these old powders burning too! Or, as you say, it may be a little over-excitment (perhaps mistaking the slight blue tinge to the methane flame for purple). In sort, go with the colour that was most prominant.
Hope this helps!
