What's The Data Look Like So Far...? One Week to End of Crowd-Source Funding Capture

 

One of the challenges in gaining support for a project like this is that it is a long-game. This isn't just limited to data collection terms, but also analysis and the changes that result as a consequence.

The current data collect has been running since 28 Dec and so far I have 711 completed surveys, and 401 'Yes' responses. I am looking for 1000 'Yes' responses before 31 Mar 14 to provide good sample sizes as I start to break the data down during the analysis process.

Below are some quick top level stats to whet your appetite from sample size of 401.

Types of Incidents

Total Out of Gas - 6%

Surfaced below agency (however defined) minimum gas quantity/pressures - 26%

Unplanned separation which led to solo ascent - 22%

Major equipment Problem - 22%

Depth distribution of sample population (diving normally takes at this depth)

0-18m - 14%

19-30m - 49%

31-40m - 21%

41-50m - 9%

51-70m - 5%

71m+ - 1%

Max Depth Planned for the Incident Dive in Question

0-18m - 20%

19-30m - 42%

31-40m - 16%

41-50m - 10%

51-70m - 7%

71m+ - 4%

Perception Errors

Misjudged Gas Consumption - 17%

Decision Errors

Error in judgement based on lack of experience - 36%

Poor decision to continue dive when should have ended it - 25%

Skills-based Error (Rules Based)

Direct Contravention of Diver Training Provided - 19%

Communication or Coordination Issues

Poor Communication - 34%

Failure to Communicate - 30%

Social Environment Factors

Direct Peer Pressure or Inferred Peer Pressure - 11%

Fiscal Pressure - 9%

Risk Perception

Did you consider the incident to be a credible risk prior to the dive? - 50% said no.

Behaviour changed since incident - 80% said yes.

Decision making

“Looks fine to me (can safely skip action or process)” - 10%

“Not really important” - 8%

“Normally ok, no need to check (don’t worry, always works out in the end)” - 12%

All of these are top level data points. The main analysis will be to split the data further into groups such as dives less than 100, less than 500. Divers aged 30 or less, Divers aged 40 or more. CCR divers. OC Recreational Divers etc. This will allow certain types of incident, factor or category to be better understood within the ‘parent’ group.

The more data that is provided, the more robust the analysis will be as I start to drill down into subsets.

If you haven’t filled in the survey, you can find it here - http://bit.ly/1ckb4fn 

If you want to support and contribute to this research project, look here www.experiment.com/garethlock