Context about collection and purpose of the dataset

Why the data was collected

Archives New Zealand regulates public offices and local authorities’ information management practice under the Public Records Act 2005 (PRA). A significant focus for us this year has been the reinstatement of our annual in-depth survey of public sector information management. Our goal in implementing the Survey of public sector information management is to collect quantitative and qualitative data about information management (IM) practice in the New Zealand public sector, i.e. public offices and local authorities. The purpose of gathering and publishing this data is also to support continuous improvement of information management practice. This survey provides baseline data for comparison in the coming years.

We will deliver the survey annually to develop longitudinal data and knowledge about how IM practice in the public sector evolves.

The survey results will:

• provide a whole-of-system view of IM in the public sector

• be able to be used by organisations as a self-assessment to benchmark their IM performance

• feed into our ongoing monitoring and reporting activities, and

• enable us to identify gaps in advice, guidance and education, and plan service delivery.

How the data was collected

The data was collected through the survey of public sector information management 2018/19 rolled-out between 17/06 and 08/07/2019. Using a direction to report (section 31 of the PRA), we mandated public offices to respond to the survey. Local authorities were invited to respond too but were not mandated to do so. We sent the survey using SurveyMonkey to 254 organisations, including 176 public offices and 78 local authorities. The survey recorded an 89.7% response rate.

We used the following acronyms in the survey:

• IM when referring to Information management or Information and records management

• PRA for the Public Records Act 2005

• IRM standard for the Information and records management standard issued under the PRA

• PO for public office

• LA for local authority Explanation of other IM terminology used in the survey can be found in our guide Key Definitions.

How the data was cleaned-up

The published dataset is a cleaned version of the data extracted from SurveyMonkey. The clean-up includes:

• Remediation of skip logic issue between two questions where survey respondents changed their responses in multiple sessions. Changing the response to the first question meant that the second question no longer applied but the response to the second question remained in the system when it should have been cancelled.

• Remediation of skip logic issue between multiple questions: due to the survey tool limitations, skip logic could not be applied to a group of questions, e.g. if response to Q12 and Q13 is ‘none’, Q14 should have been automatically skipped.

• Remediation of responses to multiple choice questions: some questions where ‘None’ / ‘Don’t know’ / ‘N/A’ options were available along with a picklist did not have any skip logic built in, i.e. respondent was able to select None and one or more options in the picklist.

• Back-coding was done if what was entered by a respondent in the ‘Other(s) (please specify)’ field matched an item from the picklist.

We removed any personal identifying information to protect the privacy of respondents. The name of the responding organisation is included.

To facilitate the use of the dataset we numbered the questions as they appeared in the online questionnaire.

Limitations

The free text option provided as the ‘Other(s) (please specify)’ field was used by some respondents to give answers not related to the question or to give more description related to their picklist option.

Documentation available

• Dataset - Formatting #1: This is the original cleaned-up version of the data extracted from SurveyMonkey. This dataset is what we used for our analysis. It does not have heading as we presented the questions in the first row and the sub-questions in the second row. Presenting the data this way allows us to filter the responses per question.

• Dataset - Formatting #2 : This is a reformatted version of Dataset #1. The columns and rows are transposed so that the columns are now question, sub question, organisation name 1, organisation name 2 with the agency’s legal name being the column header from column C onwards. This format is more programmatically-friendly. Please let us know if you find this formatting useful.

• Companion spreadsheet: We created a companion spreadsheet with the logic applied to the survey questions shown in the first sheet and the questions itemised separately on the following sheets.

• Full survey findings report: This report provides the findings of the survey, includes some observations and identifies some recommendations.