The primary purpose of our study was to select qualified items to explore and confirm construct validity at the item level in the Chinese population under the IRT and CTT frameworks. Our study contributes to the literature in both methodological and theoretical aspects.
4.1. Addressing the Methodological Issues
First, the original Tinder use motivation scale was limited by using traditional CTT and a small sample. In the current research, we advanced the research by using IRT to select qualified items using a large sample. As previously noted, IRT has several pros and overcomes some cons of CTT. It is not a surprise that we had different selected items and explored different latent structures. Specifically, item selection procedures fell under several standards, and 21 items were potential candidates for deletion, leaving 25 items for the next step of analysis. Item selection based on information functions to match a target function has advantages over CTT item selection procedures [
40]. For all 25 items, total item information, which is extremely useful in test design and evaluation, reached 16 at a wide range, meaning that it was reliable for most trait levels [
25]. Although some items were dropped, the measurement precision was kept for different levels of participants. At the time of finishing the test, it was shorter than the full-length test. Exploratory factor analysis retracted six correlated factors based on evidence from the scree plot, eigenvalues, and parallel analysis.
As noted, the original Tinder motivation use motivation scale only conducted EFA. We further advanced the research by conducting CFA and consulted multiple determining indices to confirm the structure in an independent sample. Confirmatory factor analyses provided support for the six factors. In CFI and TLI, the incremental fit indices were below the commonly accepted cutoffs [
32,
41]. However, a common criticism of incremental fit indices is that they are reliant on the badness of fit of the null model [
42]. While all correlations in a matrix are positive in a dataset, many of the item correlations are relatively low. Altogether, we viewed the results as generally supporting the six-factor structure of the dating app motivation scale without considering correlated errors.
The advanced CFA-based omega coefficients were adopted for each motivation’s reliability indicators. From this point, the present study made up for some methodological deficiencies in previous studies. For the measure of reliability, both coefficient alpha and coefficient omega were used in the present study [
43,
44,
45]. Some subscales have low coefficient alphas such as ease of communication (0.546) and fun (0.660). Various reasons may cause Cronbach’s alpha to have a low value, such as a smaller number of items. The subscale named “fun” was the case. A low value for alpha may mean that there are not enough questions on the test. That is, it may explain the low coefficient alpha of the ease of communication subscale. In contrast, CFA-based reliability estimation-omega should be the new standard in reliability estimation [
45]. In the present study, a coefficient omega of 0.671 represents the proportion of total-score variance that is due to the single factor of a subscale; that is, how reliably a total score for these three items measures the emotion-focused coping factor. For MIC, which provides an assessment of item redundancy, all values fell in the range of [0.15, 0.50], meaning that items on a factor assess the same content [
38,
39]. In the future, more reliability indices may be needed in this field [
46].
Second, the present study validated the shared motivations that were reported in the original scale, identified unique motivations among Chinese adult users, and discussed the motivations that were not found in the current sample from the perspective of cultural differences and methodology concerns. Special attention was given to the sexual ideologies of traditional Chinese culture. In particular, our results demonstrated a different factor structure from that reported in the original scale, in which a six-factor structure was supported; however, four factors were basically duplicated in the original scale. Specifically, seeking a relationship (labeled “love” in the original scale), thrill of excitement, self-worth validation, and ease of communication were identified and validated in Chinese young adults. Notably, we revised the label “love” into “seeking a relationship”, and kept the labels for the other three motivations, as indicated in the original scale. The revision of the labeling was due to the cultural factor that traditional Chinese culture does not prioritize the concept of
ai (love) in an intimate relationship [
47]. Therefore, “to find a romantic partner” or “to find someone to be with” might better reflect the desire for a relationship than for love.
The differences between the motivations found in the current results and those in the original scale also manifest in the following three ways. First, although dating apps have the reputation of encouraging users to initiate casual sex, and a handful of research has regarded such motivations as fundamental, the motivation for casual sex was not supported in the Chinese sample. A possible explanation is that the cultural difference between the two samples leads to the dissimilitude in the casual sex factor. In traditional Chinese society, sociosexual expression was considered irrational [
48]. Even though Chinese ideology has evolved to recognize that sexuality has relaxed in the past several decades, the traditional avoidance of talking about sex may still exist, and having sex with causal partners may not be considered only a moral sin, but may also be harmful to health and life. Therefore, young Chinese people may not feel comfortable talking about sex or exchanging pictures. In fact, neither “finding a one-night stand” nor “finding someone to have sex with” were identified as primarily motivated behaviors on dating apps among young Chinese users.
Second, the motivation of trendiness in the original scale was not supported in the current results. Instead, using dating apps just for fun and pleasure emerged as a main motivation. However, the “fun” motivation in the current results was related to trendiness, in the way that trendiness in the original scale was regarded as an entertainment need. One possible reason for the absence of the trendiness motivation may be that dating apps have been popular for several years, and researchers speculated that this motivation would become less strong over time when dating apps had become the established form of online dating [
5]. Given that the number of monthly active users of Momo alone exceeds 20 million [
3], we believe that this speculation was confirmed in the current sample. Third, the motivation for emotion-focused coping emerged as a unique factor that does not appear on the original scale. This motivation includes three items, reflecting that some young adults use dating apps mainly to positively improve their emotional state, relax, and reduce loneliness.
4.2. Limitation and Future Research
The present findings should be interpreted with consideration of several limitations. Firstly, although we used a relatively large, heterogeneous sample of dating app users in China, adults older than 30 were underrepresented. However, females who were over 30 years old and had not married were called “leftover women” [
49]. Such a discourse presents an increasing level of marriage pressure and anxiety among women over 30 years old. We expected female dating app users in this age group to use dating apps primarily for relationship-seeking motivations. Future research should use additional sampling methods to include a larger age span of dating app users in the sample. Secondly, our findings might not fully translate to dating app use in other countries due to cultural differences. For instance, self-worth validation motivation might be more salient among North American dating app users than Chinese users, as scholars argue that the need for high self-worth is greater in North America than in Eastern countries [
50]. Future research should explore the cultural differences in specific motivations in further detail. Thirdly, the present study did not tie specific motivations to different dating apps. Different dating apps are likely designed to target audiences with different motivations. In fact, the “all-male” Grinder app is designed for male users who seek encounters with men [
51]. Future research should link motivations with the types of dating apps among Chinese users, with particular attention paid to gay dating app use as the cultural prohibition against sexual minorities is strong in China [
52].