Multiple feature requests from SRA pilot

[koenderks]

• Allow for one decimal point in audit risk model custom percentages
• Allow for negative values in the data editor in the workflow
• Show items with negative values in the misstated items table
• Add possible ordering based on book values to selected items table
• Make the data editor smaller so that it takes up less screen
• Add option to evaluate against the overall materiality in the workflow analyses

[lottemensink]

I tested everything using BuildIt_Monetary.csv. A few remarks about the sampling workflow:

If I do the workflow with the default options including keep negative book values, and I draw a sample that includes a negative book value (seed = 866, units = items, selection = random), there seems to be a problem in the misstated items table:

It is not counted (0x) whereas it should be counted 1x I think. The evaluation table seems to be fine, it is "counted" in calculating the taint, ub, p-value, etc.

Furthermore, if I choose to execute the sample using correct/incorrect annotation (even though I provided the bookvalue), the misstated items table looks like this:

Again, the evaluation table does not seem to be affected (that is, the evaluation table does not view everything as misstated, but the misstated items table does)

Then, a general thing I noted (this is also in the module as we have it now) is that if you open one sampling workflow after another sampling workflow, you get this:

This might not be desirable if someone wants to draw multiple samples based on one file?

[lottemensink]

For the selected items table: do we only want to sort by book value? Or should auditors be able to sort based on for example ID as well? I am not sure which options would be convenient in practice.

[lottemensink]

For the overall materiality: Do we want to be able to provide some kind of evaluation table for the overall materiality as well? It looks a bit hidden in the text now. But again, I am not entirely sure what people would prefer in practice.

[koenderks]

It is not counted (0x) whereas it should be counted 1x I think. The evaluation table seems to be fine, it is "counted" in calculating the taint, ub, p-value, etc.

Don't we want to count this 0x in the statistical evaluation (the evaluation table might be wrong then)? We check them integrally right?

Again, the evaluation table does not seem to be affected (that is, the evaluation table does not view everything as misstated, but the misstated items table does)

I fixed this, it now shows:

Then, a general thing I noted (this is also in the module as we have it now) is that if you open one sampling workflow after another sampling workflow, you get this:

I know, unfortunately this is unavoidable unless there is a check in the backend of JASP if this column already exists and then automatically change the name in the created column (which I cannot do...)

For the selected items table: do we only want to sort by book value? Or should auditors be able to sort based on for example ID as well? I am not sure which options would be convenient in practice.

ID can also be a character (i.e., "hoi", "doei"), so that makes it impossible. To not keep the interface cluttered, let's keep it like this and not offer too much options.

For the overall materiality: Do we want to be able to provide some kind of evaluation table for the overall materiality as well? It looks a bit hidden in the text now. But again, I am not entirely sure what people would prefer in practice.

I added it to the evaluation table now:

and to the sampling objectives plot

Can you check again to see if these changes fix your issue(s)? If they are, please approve the changes in the review.