The dependent variables that particular script calculates are (from the script's comments):
fTE_ML = Two-error maximum length, the traditional measure of a participant's FORWARD digit span. It is the last digitspan a participant gets correct before making two consecutive errors. It is set to -1 before the start of the assessment.
fTE_TT = Two-error total trials, the number of trials before FORWARD TE_ML is reached (measure of consistency), is set to -1 before the start of the assessment.
fML = Maximal FORWARD digit span that a participant recalled correctly during all 14 trials, it is set to 0 before the start of the assessment.
fMS = The FORWARD digit span that a participant is expected to get correct 50% of all times based on overall performance during all 14 trials. It is set to 0 before the start of the assessment. MS stands for mean digit span.
bTE_ML = Two-error maximum length, the traditional measure of a participant's BACKWARDS digit span. It is the last digitspan a participant gets correct before making two consecutive errors. It is set to -1 before the start of the assessment.
bTE_TT = Two-error total trials, the number of trials before BACKWARD TE_ML is reached (measure of consistency), is set to -1 before the start of the assessment.
bML = Maximal BACKWARD digit span that a participant recalled correctly during all 14 trials. It is set to 0 before the start of the assessment.
bMS = The BACKWARD digit span that a participant is expected to get correct 50% of all times based on overall performance during all 14 trials.
The particular procedure the script implements is the one from
Woods, D.L. et al (2011). Journal of Clinicial and Experimental Neuropsychology, 33, 101-111.
That procedure differs from other, more traditional Digit Span measurement tasks you may find in the literature. Also from the script's comments:
The first time a participant makes a consecutive error, the traditional measure TE_ML is set to the last correctly recalled number of digits (e.g. if participant reaches level 8, but answers incorrectly both times, the TE_ML is set to 7). Because the digit span assessment does not exit when TE_ML is set - as is done in traditional digit span assessments- but goes on for the remainder of the 14 trials, it is possible for participants to superpass levels they previously failed (e.g participant with TE_ML = 7 may successfully recall 8 digits on a later try). The new digit span assessment measure, ML, is set to the maximal length of digits correctly recalled during 14 trials.
The particular procedure presents up to *16* digits, so it's perfectly possible to receive a score of 8, 9, 10 or even above. There are lots of intrinsic and extrinsic factors that influence memory span measurements generally (including digit span; cf.
https://en.wikipedia.org/wiki/Memory_span#Digit-span ), so those scores are not necessarily indicative of any technical issue with your script.