Learning & Innovation Research Manager| Project Management Institute (PMI)Spain
Data quality and quantity is particularly important as we think about leveraging AI on projects. Considerations include the diversity and comprehensiveness of the data that is available to us.
Have you ever encountered unexpected challenges or pitfalls while using data in your projects? How did you navigate the situation and find a resolution? Saving Changes...
Pat LidenGlobal Senior Project Manager| DBC COMPANY | FedExCaldas Da Rainha, Leiria, Portugal
I think that the big challenge for the project manager is to separate what kind of data he or she is working on. Nothing has changed since the first PMBook, except that what used to be manual and managed by the project manager, nowadays is made by the machine. What I want to say is that when we are looking for the perspective of the product(s) of the project, the concerns about data as an outcome of the product of the project always was there, the result of the business process being automated. Nowadays beyond this perspective is that project management is also being automated using Gen AI. So, the concern about data remains the same, but with more complexity considering the sophistication and data regulation's modernization and the new data generated from project management automation. Saving Changes...
Anonymous
Data quality and quantity have always been important in any project. However using data for AI modeling and for AI decision making will require project managers and teams to monitor and comprehensive assessment of the data, source, diversification and the source of AI decision making to avoid misrepresentations and errors. Saving Changes...
Shenin HassanEducation Control / Compliance Specialist | Ministry of Education UAE Abu Dhabi, United Arab Emirates
Establishing robust data collection strategies, protocols and guidelines are critical to capturing meaningful data. A common understanding (of the intend and the meaning of the data) between the data provider and data collector contributes to efficiency and error-reduction. Defining and communicating (sharable) data dictionaries would help the data provider and collector to speak the same language during the data collection process. Communication is obviously the key, however, the context of communication is more important. Saving Changes...
SHARAT CHANDRA M BTechnical Specialist| NOKIA CORPORATION INDIA LTDBengaluru, Karnataka, India
Working on the Data Analytics, we handle huge amount of qualitative and quantitative data from various customers across different locations. We have not faced such challenges as there are very robust Data handling Policies which are Categorized for access and sharing at an Organizational level. We also adhere to the strict GDPR guidelines and policies.
In case such incidents do happen, seek assistance from the Data Governance team.
Saving Changes...
Beyers StrydomMr| Cappa & D'Alberto NigeriaBellville, Western Cape, South Africa
The most import aspect is to ensure that data used are firstly relevant and then as accurate possible relating the actual field required.
This needs to be understood and also where to obtain this from and to ensure that the data obtained is relevant and accurate, both qualitative and quantitative data. Saving Changes...
Beyers StrydomMr| Cappa & D'Alberto NigeriaBellville, Western Cape, South Africa
The most import aspect is to ensure that data used are firstly relevant and then as accurate possible relating the actual field required.
This needs to be understood and also where to obtain this from and to ensure that the data obtained is relevant and accurate, both qualitative and quantitative data. Saving Changes...
Tosin IbikunleMr| Union Bank of NigeriaLekki, La, Nigeria
Feb 27, 2024 4:40 AM
Replying to Visukumar Gopal
...
When the transactions data used for analytical purpose, many times some of the irrelevant data also assumed that important for the analysis. After stakeholders clarify they don't require that information, updated version is quickly loaded in the reporting system. Due to communication gap and not understanding the needs and wants of the customer and not freezing the requirements properly, this kind of pitfalls happens. PM's should take the responsibility of this, rather than depending anyother person in the project. When PM validated and ensured the data, it made easy and over come the pitfalls.
I would also add the need for continuous engagement of all relevant stakeholders. Saving Changes...
Dean CornstubbleProject Manager| Enthalpy AnalyticalDurham, Nc, United States
From several lessons learned meetings, I found that our tracking systems had some flaws in that either the programming or the data inputs were flawed or incorrect. As a remedy for these kinds of issues creeping back into future projects, several QA checks along the way were implemented to keep those issues from occurring in the future. Saving Changes...
Sechaba KeketsiProject Officer - Learning Technologies| National University of LesothoLesotho
Data privacy on the project data where not enough was done to ensure that the data lifecycle adheres to the local data protection regulations. Saving Changes...
Mike FrenetteManager, IT PMO| Halifax Water (retired)Halifax, Nova Scotia, Canada
One of the enemies of data quality is redundancy. Organizations need to encourage single data sources so that conflicting data on the same subject can be limited. As a wise person once said, "Someone with two watches never knows what time it is." (Or something like that)
If redundancy of data is a necessiity for operational, geographic or other reasons, it must be controlled. For example, one dataset can be named the source, and access to a secondary source can warn that the data may be incorrect, referencing the primary source data. Another method is to automate replication of secondary sources so that there never is a difference, or it is only fleeting.
Often unstructured data can be a problem. Think about reports generated in your organization where there are multiple contributors, reviewers and approvers. Co-authoring tools are the best way to manage such documents so that comments and alterations can be made simultaneously in a document by many people. While these tools are now commonplace, just as commonplace are incidetns of unnecessary replication by attachments being sent around in emails, resulting in the infamous saving of documents in a personal space, and forwarding of them with the much hated "Reply All" with dates or initials built into a file name to show who changed it and when. This is a sorry practice indeed, resulting in at least as many copies of the same document as there are people in the distribution list, all of which must be combined into one document, with conflicting changes and comments addressed. What a time thief!
Co-authoring and auto versioning are our friends.
Other issues with data quality abound, but these are two that come to mind for me because of their frequency of occurrence. .
PMI Team Member
