Hi all,
I would really appreciate any help. My supervisors are not trained in ABMs so they have suggested I seek the advice of the community.
I have calibrated and frozen my model (pre-registered on OSF) and empirically validated it. To my horror I have noticed an error in my code which means my output is nonsense - I think I removed it to fix something and forgot to add it back in. In an ideal world I would like to use this mistake as an opportunity to incorporate some feedback from subject matter experts and correct some other decisions I’ve made. I would then like to re-calibrate the model and then empirically validate it again with a fresh survey sample.
However… I am running out of time in my thesis and my supervisors are questioning whether I really need to collect a fresh sample. I don’t think it’s good practice because I’ve already analysed the sample, so how can I say this won’t influence my choices? However if I do collect a new sample, I think I would have to submit 2-3 months after my funding finishes.
How much of a problem is this? I would like to publish the model in the journal of environmental psychology and my model already has some issues (like extremely patchy longitudinal data) so I wonder if this level of rigor is actually helpful. It’s intended to be an explanatory rather than predictive model too.
I’m new to the field so I don’t have much experience so it’s hard to gauge how serious this issue is. Your perspectives would be really appreciated.
Thanks,
Jesse