← CSC 316 Research Methodology in Computing
Data Collection, Ethics, and Regulatory Compliance in Kenya
Notes
Data Collection Techniques in Computing
Data collection in computer science research encompasses synthetic data generation, sensor-based logging, system telemetry, web scraping, and participant surveys. Synthetic generation creates artificial datasets when real data is scarce or proprietary. System telemetry captures operational metrics such as packet drops, cache misses, and memory footprint. When human users are involved, validated survey instruments and telemetry trackers measure behavioral patterns. Researchers must ensure that data collection methods maintain high signal-to-noise ratios and avoid sampling bias.
Data Ethics and Algorithmic Bias
Ethics in computing extends beyond academic honesty into algorithmic fairness and privacy preservation. Predictive models trained on historical data often inherit historical human biases. In facial recognition and voice processing systems, models trained on Caucasian or Western datasets exhibit drastically higher error rates when deployed in African contexts. Computing researchers bear the ethical responsibility to test for subgroup disparities, disclose model limitations, and avoid building systems that perpetuate discrimination.
Legal and Regulatory Frameworks in Kenya
Researchers conducting empirical studies in Kenya must strictly adhere to statutory bodies and legal frameworks. The Kenya Data Protection Act of 2019 governs the handling of personally identifiable information (PII). Principles of data minimization, lawful processing, purpose limitation, and storage limitation must be demonstrated in research protocols. Furthermore, academic research involving field data collection, human subjects, or deployment of novel electronic hardware requires ethical clearance from an accredited Institutional Scientific and Ethics Review Committee (ISERC) and a formal research permit from the National Commission for Science, Technology and Innovation (NACOSTI).
Plagiarism and Research Integrity
Academic integrity requires proper attribution of code libraries, mathematical formulations, and textual explanations. Incorporating third-party algorithms without citation constitutes academic fraud. Source code and datasets should be version-controlled and deposited in open repositories to uphold the scientific reproducibility principle.
Practical example
A research team deploys an Internet of Things (IoT) water leakage monitoring system in the informal settlement of Kibera, Nairobi, using ultrasonic flow sensors and GSM telemetry.
Ethical and Legal Workflow: Prior to deployment, the team obtains an ethical clearance certificate from an accredited university ISERC and acquires a research license from NACOSTI.
Compliance with Kenya Data Protection Act: Water usage patterns can reveal household occupancy and daily routines. The team strips all household phone numbers and GPS coordinates, replacing them with randomized cryptographic hash identifiers before cloud transmission.
Informed Consent: Community members sign consent forms translated into Kiswahili explaining that sensor telemetry measures volumetric pipeline pressure without monitoring domestic conversations or indoor activities.
Quiz
1. What is a major ethical concern when deploying computer vision models trained solely on foreign datasets in Kenya?
2. Which statutory body in Kenya issues formal research permits for computing field studies?
3. Under the Kenya Data Protection Act 2019, what should a researcher do to protect user location and phone numbers?
Sign in to submit and securely grade your quiz.