Creating roleplay scenarios (1) planning

These two posts are an outline of the procedure I used for creating my first e-learning roleplay scenarios for an internal client.

It’s not the last word on the subject, just a summary of what I’ve learned so far. If it's useful for anyone else, great. If anyone wants to debate any of it, that could be helpful too.

There were three clear stages:

1. planning
2. scripting
3. coding      and prototyping

This isn't about the authoring tool - we used in-house templates consisting of simple HTML and Javascript.

                                         

Stage	Procedure	Example
1.Objectives	Having identified the topic area, describe a   typical situation and the ideal behaviour in that situation.     These are the familiar ‘SMART’   objectives and it’s important that they describe behaviour rather than   knowledge.	Topic: product knowledge – differences between two types   of account     Situation: call centre – customer asks to withdraw money   with 5 days notice     Objective: learner will quickly identify whether account   is instant access or notice and, if notice, will tell customer and not   proceed with the transaction
2.Common   errors	What are the most common or most important (i.e.   damaging) errors made? The most   damaging might not be common, but need to be covered.	Examples:   Staff        ask customer what kind of account it is Where        customer insists on instant access to a notice account, staff refer to a        manager.
3.Consequences	What are the real life consequences of each   decision? When would those come to light?	Customer        doesn’t know, gets annoyed Manager        not available (delay); customer’s hopes raised then dashed; manager        annoyed at being asked when procedures are clear
4.Tracking	What do you want to track? There are two kinds of   tracking you can do with our templates:   - Actions yes/no   - subjective scores.     Each choice in a decision can increase or decrease   any score. Don’t go overboard with different – it can get quite complex especially when giving feedback. With this customer we originally collected scores for customer care and procedural   knowledge but in the end settled for one overal score for ‘performance’.	e.g. Did they ask the security questions (yes/no)?     Customer care (score)
5.Complications	What kind of complications might arise that should   be included, because they’re frequent or important?	e.g.   staff are supposed to refer a decision to a   manager but the manager is busy     a commonly used system   is down
6.Help	What kind of help should be available. There are   two kinds:     1. General   knowledge bank – this is similar to anything they’d be able to refer to in   ‘real life’. They click on an ‘information’ icon before answering the   question, but then have to conduct their own search for relevant material   2. Question-specific   help – this could be in the form of a ‘coach’ or ‘advisor’ icon, who’ll give   them a hint towards the best answer.	For these scenarios   we decided after the first prototype that the ‘advisor’ option was   spoonfeeding the trainees when the decisions weren’t really difficult enough   to merit it.     We settled for a   word-for-word reproduction of the standard product knowledge documents   available to both staff and customers, with a small amount of irrelevant   information removed.