Find the data you need

Work out what you need

Having a data plan can help you identify and address data gaps by highlighting:

• where you need to source data from
• when and how often you need to collect it
• how much data you need to collect
• whether the data you already have can answer your questions.

You need data that reflects your organisation—its goals, activities and outcomes. 

Gather data that is relevant, reliable and valid

Data can be numbers and measures or descriptions, explanations and observations. The better the quality of your data, the more useful it will be in informing good decisions.

High-quality data is:

• relevant – it directly informs the decisions that make and sustain mentally healthy workplaces
• reliable – data is accurate and complete, measures are calculated correctly, and they reflect what you need to know (not what looks good)
• valid – data actually measures what it appears to measure.

It is important for organisations to be mindful of their obligations under the Privacy Act 1988. For example, even where the organisation is dealing with existing worker information, it should not be discussed with people outside those who need to know.

Measures, indicators or metrics

It is useful to identify different types of data:

• Measures capture information about the subject of interest (e.g. measures of time and cost, frequency of claims, number of leave days etc.). Measures are objective (i.e. 2 people measuring the same attributes should get the same result).
• Indicators capture information about something that cannot be measured directly. For example, we often use indicators to measure a person’s satisfaction with work at a given time. Indicators can be subjective (i.e. influenced by people’s feelings, opinions or tastes), so they must be selected carefully to ensure they are valid and reliable.
• Metrics are calculations derived from 2 (or more) measures (e.g. ratios and percentages). They can provide useful information about the size or change in measures and indicators. An example is the change in the frequency of injury claims.

Input, output or outcome measures

Data can relate to inputs, outputs and outcomes:

• Input measures capture data about implementation (e.g. resources to implement a new strategy).
• Output measures evaluate whether the implementation process achieved its immediate aim (e.g. take up of flexible work).
• Outcome measures assess whether the strategy achieved the intended effect on broader organisational goals (e.g. number of people experiencing burnout).

Measure early signs or long-term results

Think about whether you are measuring promising, early signs of change or you want to pick up longer-term results.

• Lead indicators measure the inputs to processes and systems. They can help you to monitor resources used and identify ‘early warning signs’. Examples include staffing levels and climate scores.
• Lag indicators measure outputs and outcomes. They reflect what happened and whether your organisation achieved its goals. Examples include staff retention rates or workplace satisfaction scores.

Use existing data when you can

It is likely you already collect some of the data you need via human resources and accounting systems, such as:

• records about your people – e.g. time off, resignations, workers compensation claims
• work records and diaries – e.g. hours worked, tasks and duties
• workplace interactions – e.g. between workers or with customers and clients
• financial data – e.g. cost of injury and lost time, cost of resources for mental health.

However, raw data may not tell the whole story. Look at issues from multiple perspectives to see what the data is telling you. Sometimes relying on a single number or simple measure to represent a complex outcome or problem can be misleading. Metrics such as ratios or percentages can be useful. Supplementary measures and descriptive data can also help you to understand results.

Look at your story from different angles

• Trends – Compare results over time to understand changes over time. Do patterns emerge? Did an event or change shift workers’ outcomes or experiences?
• Data linkages – Look at ways to link data, e.g. linking wellbeing survey data with overtime costs. Or you may want to see whether internal and external factors help explain patterns or changes in human resources data.
• Different groups – Aggregated figures can mask variation between groups. Do results differ according to age, gender, team, location or occupation of respondents?
• Involving stakeholders – Look for ways to involve relevant stakeholders outside your immediate team to get their understanding of what the data is saying.

Data collection doesn’t have to be complicated or expensive. You probably already have a lot of the data and information you need.

Collect extra data when you have to

Sometimes you will have to collect additional data. These are some data sources you can use:

• Staff surveys are a low-cost way to collect worker experiences and perceptions. They can be tailored to address specific issues and you can explore issues over time. Short 'pulse' surveys may help reduce survey fatigue.
• Focus groups allow you to explore key topics in more depth.
• Worker meetings are useful for issues that affect all workers and when you want different opinions.
• Interviews (or one-to-one chats) are also useful for exploring issues in more depth, particularly sensitive topics. However, they can be time consuming and workers may worry about being identified.
• Validated tools can be simple to use because you do not need to come up with the questions. And you may be able to compare your results with industry averages. However, they may not be relevant for all contexts so make sure they meet your needs.

A final tip on selecting measures

Once you have identified the measures you want to collect, reflect back on the scope:

• How does this data support good decisions?
• Am I focused on the most important challenges or knowledge gaps?
• Are these the ‘best’ measures to meet my needs?