My first thought is that if you want to do narrative analysis, then that process should start with the data collection itself. Specifically, you would want to do as much as possible to elicit high quality stories. Of course, it can be the case that your data produce an abidance of narrative material, even if that was not your original intent.

A second thing to consider is there are a number of different approaches to narrative analysis, whereas TA tends to be highly general in its application. So, you need to be clear about which version of narrative analysis you are using.

Finally, I don't thing that the two approaches are entirely incompatible, In particular, you could do a TA to cover the board aspects of the full data set, and then pick a particularly interesting part of those results to investigate in more depth with narrative analysis.

The first difference is that a narrative approach is a methodology which incorporates epistemological and ontological assumptions whereas thematic analysis is a method or tool for decomposing data into themes, that can be used within different research perspectives.  Having said that, there are different narrative approaches, but all assume that in some way we construct meaning through telling stories.

Therefore a narrative approach would have advantages if you think that it makes sense to view what you are researching as storied in some way.  Using a narrative approach also helps to keep accounts as wholes, rather than fragmenting them into components, and means that your analysis is more sensitive to the temporal aspects of experience.  On the other hand, TA can be very useful for identifying commonality across participants, through the idea of shared themes, though your perspective on each person can become more fragmented.  TA is often seen as easier too. 

My first thought is that if you want to do narrative analysis, then that process should start with the data collection itself. Specifically, you would want to do as much as possible to elicit high quality stories. Of course, it can be the case that your data produce an abidance of narrative material, even if that was not your original intent.

A second thing to consider is there are a number of different approaches to narrative analysis, whereas TA tends to be highly general in its application. So, you need to be clear about which version of narrative analysis you are using.

Finally, I don't thing that the two approaches are entirely incompatible, In particular, you could do a TA to cover the board aspects of the full data set, and then pick a particularly interesting part of those results to investigate in more depth with narrative analysis.

I would go along with David L Morgan.

I agree with David Morgan, as I often do!  Even if you have not designed your data to elicit stories, you may find them in your data if you look for them, and they are almost always significant, revealing, whether widely shared or not.  They may provide keys that help unlock the rest of your data, suggesting deeper meaning.  And you can do thematic studies of the stories you find.  So the methods are complementary, no need ot choose either one or the other, unless your exclusive focus is on narratives and that is what you set out to collect and analyze.

irene