Tea in India was first introduced by Robert Bruce of East India Company in the eighteenth century (Sivanesan, 2013). Later, tea has been discovered by Charles Alexander Bruce in Rangpur, presently near Sivasagar district of Assam. Maniram Dewan became the first Indian to start tea plantation in Assam (Kalita, 2024). It is believed that the Singpho tribes were the first group of individuals who discovered tea centuries before Bruce found it (Sarma, 2013).
       
The tea plantation industry has always been labour-intensive, requiring a significantly larger workforce than most other sectors (Barman, 2021). But when the local people of Assam refused to work in the tea gardens, the Britishers had no other choice but to migrate workers under contractual agreements from other parts of the country such as Orissa, Jharkhand, Bihar and West Bengal (Hazarika et al., 2024). Over time, these workers settled in the tea estates forming an integral part of Assam’s demographic structure.
       
Assam’s flourishing tea sector act as a pillar of the state’s economy and remains a primary source of employment (Jain and Sharma, 2026). The industry provides employment to nearly eight lakh workers working in more than one thousand tea estates in Assam (Phukan and Chakraborty, 2025). This tea-based agro-industry ecosystem contributes substantially in sustaining the agricultural landscape and rural economy of Assam (Gogoi et al., 2016). Regardless of their crucial contribution to the state’s financial stability, many tea garden labourers are exploited to face various types of challenges like low pay, inadequate housing, lack of access to essential health and educational facilities, all of which adversely affect their QoL. The tea garden workers despite facing several hurdles resembles strength, cultural diversity and unity. There issues and challenges require immediate action and steps must be taken by government and management authorities to plan a better and secured future for the tea community.
       
Quality of Life (QoL) refers to a person’s perception regarding his or her existence and the surrounding environment in which he or she lives (Anand et al., 2023). The World Health Organization (WHO) defined QoL as “an individual’s perception regarding his/her position in life in context of culture and value system in which they live in relation to their goals, beliefs, standards and concerns” (WHO,1997).
       
There are mainly three approaches for measuring the QoL, namely objective approach, subjective approach and a combination of both objective and subjective approaches (Naess, 1999). The present study focuses on both objective and subjective approaches, as previous studies have found poor correlations between the objective factors and subjective perceptions of well-being (Cummins, 2000).
       
The objective approach measures QoL using verifiable indicators that reflects an individual’s living conditions. It includes measurable variables such as income, expenditure, literacy, housing conditions and consumption patterns (Deaton, 1997). On the other hand, the subjective approach deals with the psychological state and personal perceptions of individuals. It emphasises feelings, emotions and satisfaction levels regarding way of life and well-being (Diener, 1984).
       
In this context, the present study attempts to construct and validate composite indices that incorporate both objective and subjective perspectives for measuring the QoL of tea garden workers of Sivasagar district of Assam.

The study was conducted during the period of 2024-2025 at the Department of Statistics, Cotton University, covering 28 tea estates across the Sivasagar district of Assam. Sivasagar is recognized as one of the major tea producing districts of the state and covers an area of 2668 square kilometres. Geographically, the district shares border with Brahmaputra River to its north, Nagaland to its south, Charaideo district to its east and Jorhat district to its west.
       
The research relies on primary data collected using a structured schedule on various dimensions affecting QoL of tea garden workers by having direct and face to face interviews. The desired sample size was obtained using Cochran (1977) formula for large populations. Considering a 99% confidence interval (Z = 2.58), 5% level of precision and highest variability (p = 0.50), the sample size was estimated at 666 for both permanent and temporary tea garden workers, resulting in a final sample size of 1332 workers. A two-stage sampling procedure i.e. Proportionate Stratified Sampling and Simple Random Sampling technique has been used to ensure complete representation of workers across all the tea estates.
       
The study assessed QoL using a multidimensional schedule consisting of both objective and subjective variables. Objective variables included demographic and socio-economic characteristics like income, transport, education levels, substance consumption, nutrition and housing infrastructure. Subjective variables measured the workers satisfaction with their financial conditions, education, health-care services, living conditions, along with their psychological well-being, cultural participation, social support and personal relationships.
       
The data analysis was performed using Statistical Package for Social Science (SPSS). Principal Component Analysis (PCA) with Varimax rotation was applied separately for objective and subjective variables to identify the latent components and construct composite QoL indices. The technique focuses on extracting maximum possible variance from the dataset with each subsequent component (Tabachnick and Fidell, 2007). Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity were applied to verify whether the dataset is appropriate for factor extraction. The components with eigen value greater than one was selected following Kaiser’s Criterion, while the scree plot was used to confirm the number of components extracted. On deciding the number of components, factor loadings were examined to check the correlation between the variable and their components. Following this, Varimax rotation was applied to the dataset to reduce the number of variables with high loadings on multiple components (Tabachnick and Fidell, 2007). The resulting rotated factor loading matrix served as the foundation for further evaluation and development of composite indices. The methodology proposed by Nicoletti et al., (2000) was used for weighing and aggregation of the composite indices.
       
The reliability and validity of the constructed composite indices were evaluated using Cronbach’s Alpha, Composite Reliability (CR) and Average Variance Extracted (AVE). Cronbach’s Alpha and CR scores above 0.70 were considered satisfactory for internal consistency, while AVE scores exceeding 0.50 were deemed satisfactory for convergent validity (Cheung et al., 2023; Fornell and Larcker, 1981).

Objective QoL index
 
The suitability of objective variables for PCA was revealed by KMO measure of sampling adequacy value of 0.799 and statistically significant Bartlett test of sphericity (χ² = 15803.739, p<0.001). As indicated by Table 1 both these tests showed enough intercorrelation between the variables to perform PCA.


       
Based on Kaiser’s criterion results given in Table 2, five components whose eigenvalue were greater than one has been extracted from the objective variables. The scree plot given in Fig 1 further verifies the extraction of five components, showing a clear shift after the fifth component. Together, the five extracted and rotated components explained 75.29% of the cumulative variability considered acceptable for further analysis.




       
The structure matrix of the five components through PCA using varimax rotation is shown in Table 3. The first component explained the highest 23.35% of the total variability and was characterised by strongest factor loadings of Fruits, Meat/Fish, Egg, Vegetables and Dairy consumption and was described as Nutrition and Diet. The second component, Housing Infrastructure, explained 15.75% of the variability and involved house type, number of rooms and toilet type. The third component, Socio-Economic Status, explained 14.08% of variability and included nature of worker, monthly income and monthly expenditure. The fourth component, Education and Accessibility, explained 12.01% of variability and was represented by transport facility, electronic gadget and education qualification. Lastly, the fifth component, Substance Consumption, explained 10.08% of the variability and consisted of liquor and tobacco consumption.


       
Following Nicoletti et al., (2000) methodology, five intermediate composite indices corresponding to five extracted components were constructed using squared factor loadings scaled to unity presented in Table 4. Consequently, within the Nutrition and Diet index, the highest weights have been attributed jointly to Meat/Fish and Egg consumption (0.230), followed by Fruits (0.220), Vegetables (0.160) and Dairy (0.160). House type (0.370) contributed most to the Housing Infrastructure index, while monthly income (0.360) achieved the highest weights under Socio-Economic Status index. Electronic gadget (0.400) emerged as the most influential variable under Education and Accessibility, whereas liquor (0.540) received the highest weight under Substance Consumption. The five intermediate composite indices were aggregated by weighing each index according to its share of the total explained variance, resulting in the final Objective QoL Index.


 
Subjective QoL index
 
The suitability of the subjective variables for PCA was revealed by a KMO measure of sampling adequacy value of 0.802 and statistically significant Bartlett test of sphericity (χ² = 31234.704, p<0.001). As demonstrated in Table 5, both these tests showed enough intercorrelation among the variables to perform PCA.


       
Based on Kaiser’s criterion results in Table 6, seven components whose eigenvalue were greater than one has been extracted from the subjective variables. The scree plot given in Fig 2 further verifies the extraction of seven components by displaying a clear shift after the seventh component. Together, the seven extracted and rotated components explained 85.74% of the cumulative variability considered acceptable for further analysis.




       
The structure matrix of the seven components through PCA using varimax rotation is presented in Table 7. The first component explained the highest 15.44% of the total variability and was characterised by strongest factor loadings for physical health, life satisfaction, mental condition satisfaction and stress and was therefore described as Psychological Well-being. The second component, Financial and Social Well-being, explained 14.68% of variability and included savings, work satisfaction, salary satisfaction and socio-economic satisfaction. The third component, Living Conditions Satisfaction, explained 11.52% of variability and included housing satisfaction, sanitation satisfaction and drainage satisfaction. The fourth component, Cultural Engagement, explained 11.50% of variability and consisted of cultural connection, seasonal festivities and cultural participation. The fifth component, Social Cohesion, explained 11.06% of variability and was represented by personal relationship, care and management satisfaction. The sixth component, Health Care Satisfaction, explained 11.00% of variability and included satisfaction with health services, hospital condition and medical staff. Lastly, the seventh component, Educational Satisfaction, explained 10.55% of variability and comprised satisfaction with general education, family support and educational opportunities.


       
Following Nicoletti et al., (2000) methodology, seven intermediate composite indices corresponding to the seven extracted components were constructed using squared factor loadings scaled to unity presented in Table 8. Consequently, Within the Psychological Well-being index, the highest weights have been attributed to life satisfaction (0.270), followed by stress (0.260), physical health (0.240) and mental condition satisfaction (0.230). Financial and Social Well-being index assigned the highest weight to socio-economic satisfaction (0.280), while sanitation (0.350), seasonal festivities (0.350), personal relationship (0.360), medical staff satisfaction (0.360) and educational opportunity satisfaction (0.350) obtained the highest weights for their respective indices. The seven intermediate composite indices were aggregated by weighing each index according to its share of the total explained variance, resulting in the final Subjective QoL Index.


 
Empirical validation and comparative evaluation of the constructed QoL indices
 
To validate the explanatory power, multiple regression analysis has been used where the self-rated QoL is chosen as the response variable and the constructed intermediate indices are considered as the predictor variables.
       
As shown in Table 9, the present study shows fitted R-Square value of 0.667 in case of objective variables and 0.645 in case of subjective variables, where both cases explaining greater power as compared to Anand et al., (2023) fitted model that used WHO-QOL domains.


 
Reliability and validity of the constructed QoL indices
 
Cronbach’s Alpha and CR have been used for testing the internal consistency. AVE has been used for testing the convergent validity of the constructed intermediate composite indices.
       
From Table 10, under both approaches, all the indices have Cronbach Alpha’s value greater than 0.70 except Education and Accessibility, which is acceptable in case of exploratory research. Similarly, AVE of all the indices is above 0.50 except Education and Accessibility, but since it exceeds 0.40, it can be considered acceptable if its CR value is greater than 0.70. This is an indication of satisfying the internal consistency and convergent validity of the constructed indices.

The present study developed and validated objective and subjective composite indices for measuring the QoL of tea garden workers in Sivasagar district of Assam. Utilising the PCA technique, five objective dimensions Nutrition and Diet, Housing and Infrastructure, Socio-Economic Status, Education and Accessibility, Substance Consumption and seven subjective dimensions Psychological Well-being, Financial and Social Well-being, Living Conditions Satisfaction, Cultural Engagement, Social Cohesion, Health Care Satisfaction and Educational Satisfaction were identified and used to construct weighted QoL indices.
       
The constructed indices were found to be statistically reliable and valid with higher explanatory power than the WHOQOL-BREF based model. The results of the study highlighted that QoL among the tea garden workers is multidimensional in nature and is influenced by both observable living conditions and personal perceptions of well-being. Therefore, by incorporating context-specific indicators based on ground realities of tea garden communities, the study provides a more realistic assessment of workers QoL. Hence, these identified dimensions may be used as an input for policymakers, tea management and government agencies to design targeted interventions for improving the living and working conditions of tea garden workers in Assam.
 
Authors contribution
 
Gourab Jain: Designed the framework of the research, conducted the field work, collected and analysed the data and prepared the manuscript. Dr. Bandana Sharma: Conceptualized the research work, supervised the research process and reviewed and approved the final manuscript.

Funding
 
The authors received no external funding.
 
Ethics approval and consent to participate
 
The study was reviewed and approved by the Institutional Review Board and Ethics Committee of the concerned institution. Informed consent was obtained from all participants prior to data collection and the confidentiality and anonymity of respondents were maintained throughout the study.

The authors declare that they have no conflict of interest.