LinChiat Chang, Ph.D. is an independent consultant offering solutions in data science - including pattern discovery and feature engineering, building and deploying predictive models that learn from massive datasets, applied modeling and algorithm optimization in population health information systems and legacy big data environments, forecasting trends, classification, clustering, association rule mining, and anomaly detection using machine learning and statistics. She delivers data models and research designs informed by a strong background in social psychology and quantitative research methods such as experimental design, psychometrics, and sampling statistics.

She works with diverse organizations - from brave young start ups with fewer than 5 employees, to powerhouse foundations funding ambitious programs around the world. She assesses the validity and reliability of research findings, supports causal inference with original research designs and innovative ML algorithms, quantifies uncertainty around population projections, and reveals contingencies that limit predictive chains, as well as generalizability of observed effects to national and regional populations.

She helps develop and inform the launch of new programs and services in global health, and evaluates the impact of interventions in both proximal and long term time frames. Her research is published in peer-reviewed journals including the Public Opinion Quarterly, Psychology and Marketing, Military Psychology, Sociological Methodology, Field Methods, and more. She holds a doctorate in Psychology from Ohio State University, and did post-doctoral research at Stanford University. She founded her solo consulting practice in San Francisco, California in 2010, and is now a digital nomad based out of Cape Town, South Africa since 2020.

 

My work generally falls into one of two areas - Data and Methodology. My work in Data runs the gamut from data discovery, data evaluation, data transformation, to data analytics and visualizations grounded in statistical learning algorithms spanning both predictive and exploratory models. My services in Methodology includes survey design, sample design, experimental design, psychometrics, and quantitative aspects of program evaluation.

If you are looking for guidance on research methods or advanced analytics, I can help you. Regardless of industry sector, the same fundamental principles apply if you wish to obtain the most valid and reliable findings possible, within budgetary constraints. You can be assured at the outset that I maintain total transparency in data sourcing and modeling; so all steps will be clearly documented to withstand the scrutiny of your intended audience.

I have studied the methodological challenges underlying every step of end user research, starting from survey sample design and weighting, coverage and nonresponse errors, psychometric tool development, cognitive biases in recall and response across multiple modes of data collection, and techniques to assess data quality and veracity. I am also adept at factorial experimental designs, including discrete choice models, embedded within probability-based sample designs, thus assuring both valid causal inference as well as generalizability of research findings.

Let's talk if you want valid and reliable research output.