Data Quality Improving motivation to improve data quality

Data Quality Improving motivation to improve data quality

We need data… Reliability Growth Work with manufacturers to improve design Optimise system reliability and minimise LCC at design Add reliability & energy efficiency clauses into purchase contracts Replace at the optimum time Reduce costs by eliminating unplanned failures Optimise inspection intervals for critical equipment Reduce process variability Improve repair shop processes Provide targets for metrics Benchmarking

The “soft” bit in the middle…

Increase importance to punishments Show importance to Don’t company understand why SOLUTION…? Don’t More understand communication the about process the process it’s important Show importance to coworkers Increase importance to rewards More Don’t training on understand how to fill how in the to forms do it Set many aside specific Too other time aftertoeach things do job for data collection WHAT SHOULD YOU DO?

How to work out the best solution? Reasons Context • Which factor is important for each area? • Need a model which can incorporate & then differentiate different reasons for low DQ • How do the other goals & identities affect the DQ goal? • “Fixing the equipment is more important than collecting data”

To the safety of the equipment To my supervisor To the organization To me keeping my job Not important To how much “fun” the job is or how much control I have To my team or to a particular mate in my team To the overall maintenance of the equipment To getting more money (bonus/promoted) To how well I do my job

Goal Hierarchy Why do we do what we do? Goal choice (behaviour) driven by higher-order goals, identities & values

A Basic Hierarchy Strength of connection related to: 1. Helpfulness in achieving higher-order goal 2. Feasibility of accomplishment 3. Accountability

When “Home” is Salient

When “Work” is Salient

How Does This Help DQ? • What are the identities that people hold? • Which are most important? • Can we link ‘collecting data’ to that which is most important? – E. g. , no point in linking to “being a good employee” if that isn’t important to them • What is the feasibility of data quality? • Interventions designed around the identities and personal projects that are most common across the group

What Will We Do? • Elicit people’s goal hierarchy • Statistically link with DQ ratings to check validity of theory – Controlling for other factors such as personality, team characteristics etc • Identify common identities or projects • Design intervention based on those

What do you think? • Does this strike a chord with you? • Any factors you can think of that don’t fit this model? • Any reservations?