Best practices for TIIM

The TIIM research software is in continuous development. Researchers might design changes relevant for their research together with the BMS Lab team. This means TIIM is not static and bugs might be introduced as a result. This article describes a method of working with TIIM that can help researchers avoid potential issues both known and yet unknown.

Good working method: #

1. Familiarize yourself with the innerworkings of TIIM by going through the tutorials and this page.
   • Ensure that you have approximately 2 – 3 weeks to built your study. You might need to adapt your methodology after learning how TIIM works properly or might need time to get an answer from the BMS Lab on certain aspects of TIIM.
2. Design your study on paper (or digital paper) beforehand. What would you like to measure? When? What will it deliver afterwards? This helps your realize what you need to built, when and how much time it will likely cost. It allows you to realize any potentially impractical sides to your study and allow you to adapt them (sometimes study design has to adapt to the capabilities of the tool).
   • Take limitations of the software into account
   • Take practical aspects into account: is there an easier way to do it? Uncertain? Ask the BMS Lab.
   • Is your design feasible and/or practical for the participant?
   • Write down your measurement schedule completely
   • Try to design your study in ways that allow easy production. For example: Have recurring items? Group them in a single module that you can duplicate easily.
   • Write down your measurement items. Sometimes TIIM might offer a more intuitive or practical item type to measure something. Adapt your study accordingly.
3. Built your study in TIIM. See this as a test built, wherein you can experiment
   • Regularly test new additions you make.
   • Unsure about a feature? Check the list below or inform the BMS Lab.
   • Don’t forget to write the emails that your participants will receive and give your participants clear instructions.
   • Your introduction screen should adequately inform your participants and meet the BMS faculty’s ethical standard on informed consent.
4. Test your study thoroughly.
   • Make sure your whole study works as expected. Try breaking your conditions and routing and have some friends pilot test the study.
   • Ensure that your data output is usable. Download the CSV file (see list below on how to open this) and check whether it is what you expect and is usable for your purposes. We recommend that you duplicate your study before pilot testing it and use the duplicate for the pilot. You might have made changes when building and testing the study yourself, that could have negatively impacted the way in which your dataset is generated.
   • Are you missing data after your pilot test? Then ensure that your participants actually got the question, because in most cases this is a routing error.
   • Are you using complex timings or elaborate studies that you do not have the time to test fully (e.g.: because your study takes more than a month of daily questions), then ensure that you’re a participant of your own study and that you are a week or more ahead of your participants. This way you could potentially fix small things before your participants notice them.
5. Duplicate your study to get a ‘clean’ copy for your participants and test this duplicate once more.
6. Good luck!
   • Questions? Email the BMS Lab.
   • Found a bug or issue? Go to the troubleshooting section and fill in the form.

Hints, work arounds and specific practices: #

• The order of items within a module found in the “Modules” section and the modules within the intervention found in the “Interventions” section is the default order in which items appear and will be the order in which they appear in your dataset. You can click and drag to change this order.
• Give first and last questions of your modules recognizable names. You might otherwise have a hard time distinguishing one module from the next in your dataset. The dataset that you download at the end of your study will contain every datapoint in your intervention without a separation into modules.
• Ensure that your participants under no circumstance get the exact (!) same module more than once. You could – in theory – have multiple interventions within the same study and have the same participant take part in multiple interventions. In this case it is possible that you add the exact same module (not a duplicate) to multiple interventions. Thus, the participant can answer the same module twice. Avoid this at all costs. Duplicating that module to have a version for intervention A and intervention B is a valid strategy.
• You can download your CSV file at the end of the study. In Windows open Excel (new workbook) and go to the tab “Data”. Click “From CSV/Text” and add your dataset. This is the only way to view your data in an organized and neat way.
• Never change your study after you have started collecting data! Changing names of modules, timing rules, measurement items or even content can damage your dataset. Adding a fully new module once the study is ongoing should be fine.
• You should avoid all situations in which you have more than 150 modules in your study. This includes all interventions that are attached to your study. Dividing your study in multiple studies is possible, but not a solution. In case you expect your study to need more than 150 modules, please contact the BMS Lab. You run the risk of not being able to access your study after a memory limit has been reached.
• Always assign your participants to the intervention based on a date and not “when the intervention starts”. This will ensures greater reliability when using timing rules.
• It is possible to set a disable rule for a module. This is done in the same window wherein you set the appear conditions for a module. The disable rule will have the effect of the module only disappearing from the participant’s open modules list once that condition has been met. Meaning that to a participant it may seem that they cannot complete the module because it stays open after filling out all the items. The answers provided by the participant will however be saved once the disappear condition has been met.
• The Temporary Modules option under the “Modules” section allows you to make completing a module a time sensitive matter. The timer will start counting for each individual participant once that specific module becomes available for that specific participant. Meaning that if you get module X after you’ve completed module A and the temporary module timer is set to 2 hours, that you will have 2 hours to complete module X after you’ve completed module A.
• If you’d like to use routing, make sure that under the “Modules” section you’ve enabled this for that particular module. This option is called “Routing order” (found below “Image”) and should be set to “Specific order with routing”.
• If you’d like to have the order of your questions within your module randomized, then go the same “Routing order” option described above and set it to “Randomization per Module”.
• Thinking of reusing or reactivating an old study (more than 6 months old)? Inform what has changed in the meantime! Regular system updates mean that some more rare bugs are regularly fixed and system stability speed and stability are increased. To make use of this, you can export the module items of your study from the “Modules” and import them in a newly created study. You will have to re-do all the settings on a study and intervention level (e.g.: timings and appear conditions), but will profit from a newer built of TIIM. We strongly encourage this course of action. Be sure to also check the changelog from within the Dashboard. It is found at the bottom as “Releases”.
• Editing the CSV export of your module items (not the data, but the questions) is not possible. If you do so, you will encounter errors when importing the CSV file into a new module.