Using QR codes to collect large amounts of data from another tablet

Hi there!

We are using tablets with an app which allows teachers to monitor child development offline by recording their observations during defined activities. Once a month, a fieldworker (discusses and) collects data regarding user behaviour and child development from that tablet.

Now, it is quite a lot of data and the idea is that the app produces one (or several) QR code(s), which the fieldworker can easily collect with a KoboCollect questionnaire.

All the data would go into one (or a few) field(s) which PowerBi (connected via API) then needs to resolve into the actual fields/variables.

Couldn’t find anything by using the Search function – is there any experience with (parts of) this challenge? Is there a recommended format for the “data pile” reflected by the QR code?

Kind regards
York

When you try to use barcode or a QR-code in KoBoToolbox, you will only be able to record a single value for that particular field. Maybe this post discussed previously should give you some hint on how a barcode can be used with a pull data function to pull information in KoBoToolbox:

Thanks so much for the quick input!

If I understand correctly, the cited approach is about making use of a non-changing information (like a household number, personal ID, etc.) to pull data connected to this. What we are trying to do instead is collecting new data.

Actually, I tried it out and it is no problem to put a string of up to 4000 characters into a QR code and thus into one Kobo field, e.g.: “1-05,1-15,1-25,/75,65,70” (but I could also put a whole json-string into one QR code and field, if such a format is easier to digest for PowerBI).

So, the questions remains: Can anybody share some do’s and don’ts in this regard?

Hi @york_rff
This customization attempt is rather ambitious and interesting. I want to unpack this into various components.

I hope you mean KoBoCollect in this case. If not, kindly explain. If KoBoCollect, then you should consider the next step.

I believe this is a sufficient amount of time to have the data collected in the first step submitted to the servers; what actually prevents you from uploading the data; it is actually a whole month of waiting?

You can still submit your data collected by the teachers in advance before the health worker’s visit.

The only way to ever execute this is to ensure that all data collected by teachers is first submitted, and when you use it to pull data on the field worker’s application to pull the data.

I would strongly recommend you unpack your implementation logic and process to a more detailed approach than my attempt above. You can then use the different functionalities to execute each component. Without unpacking, it is quite difficult to really get clear support from the community.

Stephane

1 Like

Dear Stephane

Thanks for your attempt to get your head around our challenge. I am sorry that I have misled you right at the beginning.: No, the other app (on the other tablet) is not KoboCollect, but “ChildSteps child development monitor” (not relevant here, but if you are interested, you find it on the Play Store). The only relevant thing is that we want to pick data from this other app by reading a QR code produced by this app.

So, think of any QR code reflecting a number of (always different) data, organized as a series of comma separated values or as a json string. These data, once captured in a field of a Kobo questionnaire, get transferred to PowerBi (via API) in our case, but probably there are general aspects to be observed, independently from how and where the data get transferred for processing?

Kind regards
York

Hi @york_rff
Do you have a sample of a generated QR code; you can share it privately if it has personal identifier information
Stephane

1 Like

Hi Stephane

No sample yet, because the experiences we hope to find are meant to help us to take a decision on how to structure the data and which format to use (e.g. series of comma-separated values in the fixed order of predefined positions or json key/value pairs).

Basically it are numeric values, representing attendance numbers for up to 36 standard dates in an academic year, enrolment numbers for 6 standard dates, percentage of children with the highest and lowest rating/”performance" level, number of ratings recorded, etc.). So, not very special data, but many, which means that we need a simple and flexible enough approach to handle them in the data analysis.