Add to retrospective: How many hours did you work on this project?

[jeffreywescott]

@shereefb commented on Wed Jul 13 2016

Tracking hours worked per cycle per project so we can use it in the future to calculate XP (instead of cycle length)

---

Proposed language (from @tannerwelsh):

How many hours were you actively learning and building? Consider total active hours for the cycle, and subtract any time missed.

[ Input: numeric, allowing any integer greater than 1 ]

How much time did you dedicate to each project? These must add up to the total given above.

• #project-1: [ Input: numeric, allowing any integer greater than 1 ]
• ...
• #project-n: [ Input: numeric, allowing any integer greater than 1 ]

[shereefb]

@jeffreywescott can we move this to top of backlog. It would be great to have it in.

[jeffreywescott]

There was only one thing ahead of it, which is related. To support this, we need a "self" response type, which is a bit more work. I think we can get it in place for the next cycle.

[bundacia]

@shereefb: Should this ask for a percentage instead of hours? I'm worried about this scenario. Say we have a team of three people Amy, Bob, and Cheryl. Amy and Bob didn't take any time off this week, Cheryl was out Friday. When they get to this question on the survey they all give different answers:

• Amy thinks, "I was here from 8:30-6:30 every day, so that's 10 hours a day for 5 days," and logs 50 hours
• Bob thinks, "I was here from 8:30-6:30 every day, and we usually take about 30 minutes for lunch, plus about 30 minutes for other miscellaneous breaks, so that's 9 hours a day for 5 days,". Bob logs 45 hours
• Cheryl thinks, "A work week is 40 hours, and I missed 1 day" and logs 32 hours.

The data we need is Amy = 100%, Bob = 100%, Cheryl = 80%, but there's no way to glean that from these responses. Asking this as an "hours" introduces a lot of ambiguity.

Maybe we should word the question "What percentage of this week were you absent?". Or something like that.

[shereefb]

@bundacia I think I understand what you're solving for, but hours feels a lot more concrete than percentage of a cycle. We can solve for this by clarifying the question itself, but for now I think it's a good start.

In the future, I can imagine there will be balancing stats (for example a velocity stat: XP gained per hour, and total hours worked) that will incentivize players to be honest about what they report.

[bundacia]

I think we can get total hours without making the player report them. Since all players are not supposed to do any work outside of the official hours we know how many hours are in every cycle. We don't have to rely on users to report the total, we can have the moderator set that. If you really want users to report hours and not a percentage maybe it would make sense to ask how many hours they were absent, rather than how many they worked. The we can subtract that from the known # of hours in the cycle.

My concern isn't really that hours are less precise that a percentage, just that different people will have different ideas of how long a cycle is in hours, which could lead to bad or inconsistent data.

[jeffreywescott]

@bundacia -- how would we handle the person who had a dentist appointment? Or was sick for two days?

I understand your concern about the reporting differences, but we can look at how people report the data and see how varied the reporting is, no? It seems like hours tracking is something that people will commonly have to do (for example, with client work). We could even recommend that they try out something like RescueTime or similar tools. WDYT?

[bundacia]

@bundacia -- how would we handle the person who had a dentist appointment? Or was sick for two days?

@jeffreywescott: which strategy are you referring to? If they're reporting a percentage that seem pretty straight forward. If you were out 2 days you worked 60% of the cycle. If you were out half a day for a dentist appointment you worked 90% of the cycle. Similarly, if they're reporting hours absent, the question can just prompt them to consider a full day 9 hours (or 8, or whatever). So if you're out two days you were absent 18 hours, if you had a 3 hour dentist appointment you were absent 3 hours.

I think this conversation is outgrowing the medium here. Maybe we can have a requirements discussion in a zoom call around this before implementation.

[jeffreywescott]

I'm talking about hours.

From a user perspective, calculating what percent of the time I was gone is way, WAY harder than remembering I had a 2 hour dentist appointment.

[bundacia]

Yeah, I agree with that. My primary tension is that "How many hours did you work on this project?" is very open to interpretation and isn't going to give us consistent data. I think asking for a percentage or hours absent is going to give us more consistent results. The one problem with "hours absent" is that if a player worked on multiple projects we still don't know how much time they spent on each. And as you mentioned, percentages are awkward. Maybe what we need is something like this.

How did you spend your time this cycle:
Hours spent on #fluffy-bunny: ____
Hours spent on #angry-cockroach: ____
Hours absent: ____

Your responses must add up to 40 hours.

This still leaves the issue of how they report hours spent on process things like home groups and lunch & learns etc.

[shereefb]

blocks #150

[jeffreywescott]

For what it's worth, I wasn't suggesting we capture "hours absent". I was just pointing out that if I were absent for 3.5 hours, it's easier for me to say, "I worked 36.5 hours on this this week." than figuring out what percentage of 40 hours 3.5 is and reporting 92%.

I think we could just as easily resolve @bundacia's tension by having the maximum hours be 40 (across all projects).

[jeffreywescott]

@shereefb -- quick note -- this doesn't really block #150 -- we could still show people stats, it's just that the stats will be based on an assumption that people worked (40 / n) hours on each project, where n is the number of projects they were on.

[jeffreywescott]

Reword: How much time did you spend moving this project forward (or leveling yourself up to help to be able to move the project forward)? Don't count time you were at lunch, home group, or playing Pokemon GO.

@tannerwelsh -- from UX role, please come up with good wording for this question.

[bluemihai]

FWIW, I'm finding that a lot of the work I'm doing this week especially is relevant to BOTH projects, so I can imagine being truthful and having the numbers add up to more than 40 hours.

[jeffreywescott]

@bluemihai -- that doesn't make sense. Just because work that you're doing applies equally to both projects doesn't mean you worked more hours. If it applies to both equally, it seems to me just allocate 1/2 the hours to each project. If it applies to one more than the other, perhaps 1/3 vs. 2/3 of the hours.

The whole point of capturing this is so that we can calculate how many effective hours you've worked, NOT how many hours of value you've contributed to a team.

[tannerwelsh]

@jeffreywescott added question wording to the issue description.

[bundacia]

This will replace the work-around question added as part of #346

[jeffreywescott]

@bundacia -- do we need this issue any more? I get that #346 was a different implementation strategy, but it still represents this unit of work, right?

[jeffreywescott]

I guess /cc @tannerwelsh because the way it's implemented now doesn't fully meet the specs here.

[tannerwelsh]

Moved the advanced version here: https://github.com/LearnersGuild/game/issues/361, @bundacia feel free to close this.