Dependent variable: considerations for online potted plant purchases.

The survey instrument was developed based on a preliminary survey. It encompassed 47 self-designed items related to the factors considered in online potted plant purchases. The preliminary survey was conducted with online plant-related communities to identify potential items related to online potted plant purchase considerations. Participants were administered an open-ended questionnaire asking them what features they find attractive when buying potted plants online and why those features are important. According to Taiwan’s Civil Code and academic ethics reviews, the majority age in Taiwan is 20 years; therefore, participants were individuals older than 20 years with online shopping experience. The sample size determination adhered to the concept of theoretical saturation, which implies that data collection can stop when additional surveys with new samples yield no new information (Glaser and Strauss 2017). The survey, conducted from 9 to 26 Aug 2021, in Taiwan, garnered 83 valid responses from two large online plant communities, namely Moss Lovers Micro Landscapes and Ferns Fantasyland. As no new information was found, the data collection was stopped.

Three researchers with horticultural backgrounds categorized the survey data and identified common concepts independently, and then discussed and reached a consensus. The interrater reliability was 97%, indicating a high consensus. The initial list of 47 items related to online potted plant purchase considerations (Table 1) covered plants, products, functions, online shopping, shipping, and emotion.

Table 1. Initial list of 47 items related to online potted plant purchase considerations based on preliminary survey.

View Fullscreen

The 47 items were used in the formal survey. The survey statement was, “Please assess the importance of the following items in influencing your willingness to purchase potted plants online.” The study used a 7-point scale to assess the importance of each item, ranging from 1 (not extremely important) to 7 (extremely important).

Independent variable: personal factors.

The survey instrument encompassed seven items about personal background (gender, age, educational level, monthly income, marital status, children, and children living at home), five items about the living environment (perceived naturalness, building density, park abundance, outdoor space, and indoor space for plants), and four items about previous experience (gardening education, online plant, plant materials, and online nonplant product purchases). In the personal background section of the survey, participants were asked to disclose their gender, age, educational level, and personal monthly income. Age was categorized into 5-year intervals, ranging from 20 to 65 years. The categorized educational levels ranged from primary school to postgraduate education. Following the previous studies of Zhao et al. (2016), the survey collected information on marital status and the status of any children in the home. Marital status was categorized into married and other options. Children's status asks participants the number of children they had in total and how many of those were living at home.

The living environment influences plant purchases, as evidenced by variations between urban and rural areas (Yeh and Huang 2009) and regional differences (Campbell and Hall 2010). This study adopted Hur et al.’s (2010) concept of the living environment, which focuses on naturalness, vegetation rate, openness, and building density. The survey assessed participants’ living environment through three questions on the perceived naturalness, perceived building density, and perceived park abundance of their living environment. Furthermore, having more living space at home increases the likelihood an individual will purchase plants (Zhao et al. 2016). Specifically, the home environment influences potted plant consumers’ choices. The questionnaire included items related to the home environment about participants’ outdoor (e.g., yards, balconies, terraces) and indoor (e.g., entryways, living rooms, bathrooms, bedrooms, and studies) space for plants.

Previous research suggests that gardening education may influence consumers’ willingness to purchase potted plants (Phillips et al. 2007). This survey measured gardening education by asking respondents if they had received horticulture or life sciences education. Individuals with experience in online shopping tend to hold more favorable opinions regarding security, return policies, and trust in online retailers, perceive lower risks, and show a greater propensity to make online purchases (Lin et al. 2010). Thus, the survey also asked participants about the frequency of their online purchases in the past year, including plants, plant materials, and nonplant products. The study used a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree) to evaluate participants’ living environment and previous experience with plants and online shopping.

Data collection and analysis.

The Ethics Committee of National Cheng Kung University reviewed and approved this study (approval number: NCKU HREC-E-110-139). The research used age quota sampling to target individuals older than 20 years with experience purchasing plants online. The age groups were 20–29, 30–39, 40–49, and 50 years or older. This study was conducted at the Taiwan Foliage Plant Exchange Station, which is Taiwan’s largest online plant community, from 14 to 16 Mar 2022.

The study conducted an exploratory factor analysis (EFA) on 47 items related to online potted plant purchases to identify common factors. Following Williams et al. (2010), the study applied principal components analysis with varimax rotation, focusing on factors with eigenvalues of 1.00 or higher. The Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test evaluated the suitability of the data for factor analysis. The factors were refined by applying multiple criteria for item reduction, including removing items with low factor loadings and high cross-loadings, evaluating internal consistency using Cronbach’s alpha, identifying item-to-total and interitem correlations, and ensuring a minimum of three items per factor for meaningful analysis (Earl 2012; Hair et al. 2019). Harman’s single-factor test assessed the influence of common method variance, using the unrotated factor solution from the EFA (Chang et al. 2010; Podsakoff 2003).

The study used multiple regression analysis to examine the effects of personal factors on these common considerations in online potted plant purchasing behaviors. For non-normal data, bootstrapping with 5000 resamples was used to calculate confidence intervals. The collected personal data were transformed into dummy variables to perform a multiple regression analysis.

Descriptive statistics.

A total of 400 valid questionnaires were collected, with each age group contributing ∼100 responses (Table 2). Women constituted most of the participants (63.7%). The most common educational level was university (51.7%). The largest income bracket among participants (48.5%) was NT$25,001–50,000 (USD1 = 30 NT). Regarding marital status, nearly half of the respondents were married (48.5%). Most participants (60.0%) did not have children; however, of the 40% who did, most had children living at home.

Table 2. Sociodemographic characteristics of 400 respondents surveyed in Taiwan’s largest online plant communities through age quota sampling from 14 to 16 Mar 2022.

View Fullscreen

Online potted plant purchase considerations.

The initial EFA yielded a KMO value of 0.88 and a significant Bartlett's test (χ² = 5405.51, df = 666, P < 0.001), confirming the appropriateness for factor analysis. The EFA extracted six common factors comprising 21 items (Table 3). These factors were labeled consumer assurance, plant care and safety, well-being, service quality, pricing and promotions, and size and habits. Factor loadings ranged from 0.52 to 0.84, indicating strong correlations among the items in each factor. Despite cross-loadings of 0.41, 0.44, and 0.45 on three items, the six-factor structure remained distinct. Reliability coefficients for each factor varied from 0.71 to 0.80, signifying good reliability. The final EFA revealed a high KMO value of 0.86 and a significant Bartlett's test (χ² = 17,192,563.62, df = 210, P < 0.001). The six factors accounted for 64.28% of the total variance, indicating practical significance (Hair et al. 2019). Harman’s single-factor test showed a 30.52% unrotated factor variance, indicating that common method variance had a minimal impact.

Table 3. Exploratory factor analysis of common factors considered when purchasing potted plants online (well-being, consumer assurance, service quality, pricing and promotions, size and habits, and plant care and safety).

View Fullscreen

Well-being, which is primarily related to healing, stress relief, and positive emotions, is the most important factor affecting online plant purchases. Consumer assurance, the second-ranking factor, encompasses personal data security, consumer protection, risks in consumer behavior, and company reputation. The third crucial factor, service quality, prioritizes damage-free shipping, followed by fast shipping, product descriptions, and after-sales service. The fourth key factor, pricing and promotions, emphasizes products’ shipping suitability and reasonable pricing, including shipping costs. The final two factors focus on plant-specific attributes, size, and habits, which cover plant and product size, plant habits, and plant care and safety, emphasizing nonallergenic and nontoxic plants, as well as easy plant care.

This study considered various factors such as plant quality, container color and material, photo preferences, and plant functionality; however, these were ultimately excluded during the EFA process. The results identified several common factors considered when purchasing potted plants online, including well-being, consumer assurance, service quality, pricing and promotions, size and habits, and plant care and safety (Table 3). These findings align with previous research by Yeh and Huang (2009), which identified emotional conditioning as a key motivator for purchasing potted plants. A contribution of this study is that well-being was identified as the most important and common factor in online potted plant purchases (variance explained = 12.33% in Table 3). Consumers prioritize viewing plant images online to evaluate plants’ potential to offer healing, stress relief, and positive emotions (Table 3), thereby enhancing their well-being through their potted plant purchases.

Past studies on online purchase behavior have focused on service quality and trust (Gefen et al. 2003; Zeithaml et al. 2002). In the context of online plant purchasing, consumer assurance and service quality should align with the dimensions of service quality (i.e., assurance, reliability, and responsiveness) as outlined by Parasuraman et al. (1985). Key considerations include data security, consumer protection, company reputation, fast and damage-free shipping, and effective after-sales service (Table 3). Detailed product descriptions also enhance perceived reliability. Moreover, pricing and promotions correspond with consumer attraction to reasonable prices and sales promotions.

This study identified product and plant size, plant habits, and ease of care, including nonallergenic and nontoxic qualities, as important factors (Table 3). However, plant care (Mean = 5.30) and plant size and habits (Mean = 5.47) are relatively less critical compared with consumer assurance (Mean = 6.33), service quality (Mean = 6.08), and pricing and promotions (Mean = 5.60) (Table 3). This may be because those who purchase potted plants online often have experience with similar plants or gardening, which diminishes their concerns regarding plant-related issues when making these purchases.

Effects of personal factors on online potted plant purchase considerations.

The study aimed to increase the current understanding of how personal factors influence six common considerations consumers have when deciding whether to purchase plants online. Both Kolmogorov-Smirnov and Shapiro-Wilk tests yielded significant results, indicating a non-normal data distribution. Consequently, bootstrap resampling was performed 5000 times to estimate the standard error, confidence intervals, and P values for the multiple regression analysis. This approach sought to provide more robust and reliable results, particularly given the non-normal data structure observed in the study. All the analyses revealed variance inflation factor values below 3.0, confirming the absence of collinearity issues among the variables and ensuring the validity of the regression results.

Overall, personal factors significantly affected well-being (F = 4.07; P < 0.001; adjusted R² = 0.11), consumer assurance (F = 4.02; P < 0.001; adjusted R² = 0.11), service quality (F = 4.07; P < 0.001; adjusted R² = 0.08), pricing and promotions (F = 2.76; P < 0.001; adjusted R² = 0.07), plant care and safety (F = 8.77; P < 0.001; adjusted R² = 0.24), and size and habits (F = 3.33; P < 0.001; adjusted R² = 0.09) (Table 4).

Table 4. Effects of personal background, living environment, and previous experience on six common considerations in online potted plant purchasing behaviors.

View Fullscreen

Women (B = 0.24; P < 0.01), individuals with lower educational levels (B = –0.09; P < 0.05), and those living in environments with higher perceived naturalness (B = 0.07; P < 0.01), greater building density (B = 0.06; P < 0.01), and an abundance of parks (B = 0.06; P < 0.05) emphasized well-being more when purchasing plants (Table 4). Women (B = 0.23; P < 0.05), individuals with living environments that had a higher perceived abundance of parks (B = 0.06; P < 0.05), and those who frequently purchased nonplant products online (B = 0.07; P < 0.05) were more focused on consumer assurance. Individuals with less gardening education (B = –0.06; P < 0.05) and those living in environments with higher naturalness (B = 0.09; P < 0.01) and building density (B = 0.08; P < 0.01) placed higher importance on service quality. People in living environments with higher naturalness (B = 0.10; P < 0.01) and building density (B = 0.08; P < 0.05) and those who frequently purchased nonplant products online (B = 0.11; P < 0.01) were more concerned with pricing and promotions.

Women (B = 0.38; P < 0.05), older individuals (B = 0.07; P < 0.05), and those with primary or secondary educational levels (B = –0.15; P < 0.05) ascribed higher importance to plant care and safety (Table 4). People in living environments with higher perceived naturalness (B = 0.17; P < 0.001) and building density (B = 0.16; P < 0.001) also demonstrated a greater emphasis on plant care and safety. Similarly, a lower frequency of online plant purchases indicated a stronger focus on plant care and safety (B = –0.16; P < 0.001). Women (B = 0.39; P < 0.01) and individuals with higher income (B = 0.14; P < 0.05) placed more importance on plant size and habits. In addition, having a living environment with greater perceived naturalness (B = 0.11; P < 0.05) and higher building density (B = 0.17; P < 0.001) was associated with a stronger focus on size and habits.

Previous studies indicate that women are the primary consumers of potted plants and flowers, and often search for information related to gardening online (Behe et al. 2013; Mason et al. 2008; Phillips et al. 2007). Conversely, men show higher intentions to purchase gardening products online (Behe et al. 2013). This study suggests that women place more emphasis on well-being (B = 0.24; P < 0.01), consumer assurance (B = 0.23; P < 0.05), size and habits (B = 0.39; P < 0.01), and plant care and safety (B = 0.38; P < 0.05) when purchasing plants online (Table 4). Women also tend to focus on the emotional enhancement and stress relief plants can provide (Mason et al. 2008), which leads them to value the benefits potted plants offer for well-being. Moreover, women typically buy flowering plants to decorate their homes, whereas men usually purchase them as gifts (Yue and Behe 2010). This variance in purchasing objectives may account for why women pay closer attention to plant size, habits, and care, as these factors reflect their use of plants in home decoration and involvement in plant care and safety. Satisfying the considerations this study’s findings highlight for female consumers could enhance their intentions to purchase plants online.

This study suggests that older individuals prioritize plant care and safety in online plant purchases (B = 0.07; P < 0.05) (Table 4), likely because of having children or pets in the home, which raises concerns regarding allergenic or toxic plants. Highlighting that plants are easy to maintain, allergen-free, and nontoxic may enhance the willingness of older adult consumers to buy them. Educational levels ranging from primary to high school were associated with a stronger emphasis on well-being (B = –0.09; P < 0.05) and plant care and safety (B = –0.15; P < 0.05) (Table 4), which might be associated with age. Notably, higher personal income was found to correlate with greater concern for size and habits (B = 0.14; P < 0.05) (Table 4). Individuals with higher incomes, who potentially have large living spaces, gardens, or balconies, may prioritize different plant attributes than people with lower incomes, who may contend with space limitations. Income notwithstanding, both lower and higher income groups value well-being (B = 0.01; P > 0.05), consumer assurance (B = 0.06; P > 0.05), service quality (B = 0.01; P > 0.05), pricing and promotions (B = –0.06; P > 0.05), and plant care and safety (B = 0.09; P > 0.05) (Table 4). Although some research, such as Zhao et al. (2016), has shown that larger families or those with children often purchase fewer fresh flowers, this study found no relationship between marital status or having children and the six factors that commonly influence online plant purchases.

The study found that consumers in areas with high perceived naturalness prioritize well-being (B = 0.07; P < 0.01), service quality (B = 0.09; P < 0.01), pricing and promotions (B = 0.10; P < 0.01), size and habits (B = 0.17; P < 0.001), and plant care and safety (B = 0.11; P < 0.05) in online plant purchases (Table 4). The building density produced the same results. The results align with those of past studies suggesting that urbanites have higher intentions to purchase plants (Behe et al. 2013). In areas of high building density, residents seek emotional improvement, which enhances the demand for and intentions toward potted plants. Similarly, those in more natural environments, who may have greater intentions for plants and gardening, emphasize these considerations. A higher perceived abundance of parks also correlates with a greater emphasis on well-being in plant-purchasing decisions. However, the availability of outdoor and indoor spaces for plants does not influence online plant purchase considerations (Table 4), suggesting that housing space may not be a significant factor. Overall, living in dense urban areas with natural spaces increases the importance consumers place on common considerations in online plant purchases.

Previous studies suggest that individuals with horticultural experience tend to design or produce their own potted plants (Phillips et al. 2007). This study found that people with less gardening education emphasize service quality more highly (B = –0.06; P < 0.05) (Table 4). This likely stems from their limited experience in plant care, species knowledge, and design ideas, which leads them to seek comprehensive service and explanations from sellers. As a result, most potted plant consumers may lack this background knowledge, thus making service quality an important factor in their purchasing decisions. The study also revealed that individuals who frequently purchase nonplant products online are more focused on consumer assurance (B = 0.07; P < 0.05) and pricing and promotions (B = 0.11; P < 0.01) (Table 4). Furthermore, those who lacked experience with purchasing plants online showed greater concern for plant care and safety (B = –0.16; P < 0.001) (Table 4). This may reflect increased caution regarding plant care and safety due to inexperience with online plant purchases.

Limitations of the study and recommendations for future research.

Two measurement issues warrant further discussion. The choice between using a balanced or unbalanced scale considers the nature of the issue and the distribution of past variable data. This study inducted the items through a preliminary survey. However, the importance of these items has not yet been assessed. Thus, in the formal surveys, this study employed a commonly used balanced 7-point Likert scale. The consistent use of the same balanced 7-point Likert scale also helps reduce the cognitive burden on respondents (Liao 2014; Schaeffer and Presser 2003). Ultimately, the data extraction of the factors and items revealed a negatively skewed distribution of responses. If variable data exhibit a positively or negatively skewed distribution of responses, using an unbalanced scale can be meaningful, for instance, for measuring well-being (Liao 2014). However, the difference between balanced and unbalanced scales remains unresolved and may be issue-dependent (Liao 2014). Future studies could use an unbalanced scale for factors that consider online marketing strategies to achieve a more precise measurement scale.

Various methods exist for measuring the age variable in horticultural issues, including generational cohorts (e.g., Campbell and Campbell 2019), age intervals (e.g., Yue and Behe 2008, 2010; Zhao et al. 2016), and specific age years (e.g., Behe et al. 2013; Mason et al. 2008). Generational cohorts often share similar attitudes, values, preferences, shopping habits, and behaviors (Eger et al. 2021). The COVID-19 pandemic has influenced the shopping behavior of the oldest generation, resulting in decreased online purchases and increased acquisition of food and medicines from traditional shops, in contrast to younger generations, primarily for health and stress reasons (Eger et al. 2021). The COVID-19 pandemic has introduced complexities to the issues of online consumption in this study. As a result, we chose 5-year age intervals for our analysis, focusing specifically on age rather than generational differences. Future studies should include generational cohorts post-COVID-19 to delve deeper into their effects on online plant consumption.