DATA QUALITY AND TRUST IN MARKETING AND SOCIOLOGICAL RESEARCH: IMPLICATIONS FOR BUSINESS AND CIVIL SOCIETY ORGANISATIONS IN THE CONTEXT OF WAR AND ARTIFICIAL INTELLIGENCE

Abstract

Introduction. The full-scale war has changed the conditions under which Ukrainian businesses, civil society organisations and donor programmes obtain knowledge about markets, communities and vulnerable groups. Migration, the occupation of parts of the territory, unstable communications, security risks, respondent fatigue and the social sensitivity of research topics complicate sampling and the interpretation of findings. At the same time, research practice is increasingly moving towards digital tools and artificial intelligence, which accelerates data processing but also increases the risks associated with bots, synthetic responses, automated conclusions and the erosion of trust.

Aim. The aim of the article is to substantiate the role of data quality and trust in marketing and sociological research under conditions of war and the use of AI, and to develop an applied model for businesses, civil society organisations, donors and research agencies.

Materials and Methods. The methodological basis of the article combines conceptual analysis, comparative generalisation, normative and ethical interpretation, and authorial modelling. The problem is examined through the interrelation of several dimensions: wartime constraints on data collection, risks related to the use of AI, professional requirements for the quality of marketing and sociological research, and the practical needs of businesses, civil society organisations and donor programmes.

Results. Data quality is interpreted as a set of characteristics that includes sample validity, audience reachability, authenticity of responses, fraud control, ethical data collection, privacy protection, methodological transparency and the correctness of interpretation. For businesses, reliable data reduce the risk of errors in demand forecasting, pricing, assortment planning and communications. For civil society organisations, such data help to assess more accurately the needs of internally displaced persons, veterans, communities and people with traumatic experiences. The author’s model brings together method selection, response verification, human oversight of AI-generated outputs, mixed methods and professional standards.

Prospects. Further research should focus on testing the proposed model in business, humanitarian and territorial projects, as well as on developing methodological profiles for studies related to community recovery, the development of territorial and local brands, and the responsible use of AI.