Crowdsourcing for widening participation and learning opportunities: a view from language learners’ window

Online crowdsourcing sites/platforms have become popular in recent years. This study aims to uncover when, where, and how language learners in Turkey (TUR), Poland (POL), Macedonia (MAC), and Bosnia and Herzegovina (B&H) make use of the available crowdsourcing websites/games to learn foreign languages. To ensure parallelism among the data collected in the four countries, a cross-culturally appropriate online questionnaire in English comprised of two parts was designed for this study. Part one gathered information about the use of crowdsourcing sites, tools, and games, while part two elicited background information related to the participants (N=211). The data were analysed considering countryand contextspecific variables. The results show that there are more similarities than differences in the ways informants in the studied countries perceive, and employ crowdsourcing resources to learn languages. Therefore, the findings might provide insights for experts, material developers and teacher trainers striving to create cross-culturally valid crowdsourcing platforms/games.


Introduction
Online crowdsourcing sites/platforms that depend on the contributions of ordinary users for their development and growth have become popular in recent years (e.g. Wikipedia, Busuu). This popularity led to an increase in the number of foreign Language Learners (LL) using them, as they provide easy and free access to engaging culture-and context-specific materials. Little is known, however, about how LL with different linguistic and cultural backgrounds view the available online crowdsourcing sites/games, or how they employ them. Therefore, in this study, we focus on four linguistically and culturally diverse countries -TUR, POL, MAC, B&H -and aim to uncover when, where, why, and how LL make use of crowdsourcing websites, tools, and games to learn Foreign Languages (FL). We believe that the findings of the study could provide valuable insights to experts, material developers, and teacher trainers striving to create crowdsourcing platforms that are valid across cultures.

Data collection
To ensure parallelism among the data collected in all countries, a cross-culturally appropriate questionnaire (i.e. a tool that was free of culture bias, comprehensible, and relevant to all participants) was designed for this study. The data-gathering tool was in English, and comprised of two parts. "Part A: crowdsourcing" included 11 checkbox, Likert scale, and open-ended questions eliciting data related to the crowdsourcing practices of the participants (i.e. how, when, where, and why LL use various crowdsourcing platforms). In "Part B: background information", there were four checkbox and two open-ended questions eliciting information related to the participants.

Data analysis
The data collected from online surveys were thematically classified and analysed with descriptive statistics considering country-and context-specific variables.

Results and discussions
This study focused on crowdsourcing for widening participation and LL opportunities. Therefore, our first objective was to uncover how participants conceptualised crowdsourcing. As such, Item 1 in the questionnaire was What comes to your mind when you see/hear the word crowdsourcing?. Despite being a relatively recent concept, only three out of the 211 participants said that they "have never heard that word" (see Figure 1). The remaining 98.6% knew what crowdsourcing was, and despite the differences in the frequency with which it was selected, Definition 2 was the most popular in all countries; 65.9% of the participants thought that crowdsourcing was "a 5. Common European Framework of Reference model where information is gathered from different people". Definition 4, which highlights that the contributors to "the crowdsourcing activity might not be experts in the field", was a distant second (22.3%) choice. Finally, Definition 1, where the division of labour "among the participants to achieve a cumulative result" is emphasised, was participants' third choice.
One explanation for the lack of consensus among the participants in choosing the definition of crowdsourcing (TUR=85.4%>B&H=71%>POL=60.3%>M AC=48.8%), comes from Estellés-Arolas and González-Ladrón-De-Guevara (2012), who argue that the term crowdsourcing "encompasses many practices. This diversity leads to the blurring of the limits of crowdsourcing that may be identified virtually with any type of internet-based collaborative activity" (p. 189).
Regardless of the differences, our participants mostly selected Definition 2. Why? One reason for this could be the word itself. Crowdsourcing is "formed from two words: crowd, making reference to the people who participate in the initiatives; and sourcing, which refers to a number of procurement practices aimed at finding, evaluating, and engaging suppliers of goods and services" (Estellés-Arolas & González-Ladrón-De-Guevara, 2012, p. 89).
So, opposite to Howe's (2006) claims, it looks as if the crowd from whom the data in this study were collected (LL) used their knowledge of etymology to select the definition of crowdsourcing.
The next question in the survey was related to the crowdsourcing sites/tools participants were using to learn FL. Wikipedia was the most popular site in TUR, MAC, and B&H, and a close second in POL (Figure 2). Other frequently used sites/tools were Kahoot and Duolingo. Bergvall-Kåreborn and Howcroft (2014) argue that crowdsourcing platforms tapping into concepts such as "collaborative consumption, community building, the sharing economy, and social enterprise" (p. 215) become popular with users and contributors. For the participants in our study, Wikipedia was the crowdsourcing site that ticked all of these boxes.
When the participants were asked to identify the games and devices they utilise to learn FL, we found results supporting Blume (2020), who reported that LL do not make much use of games while learning FL. Overall, 80.6% of our informants stated that they had not used any games to learn FL, with slight differences between the countries (POL=86.2%>TUR=83.7%>B&H=79.7%>MAC=70.7%). In regard to the utilised devices, 93.8% of the informants selected smartphones, and 85.8% laptops. About one-third (28%) stated that they use personal computers, and a relatively small number identified tablets (19%) and iPods/iPads (12.8%) as devices they employ while learning FL. Figure 2. Answers to "Tick the crowdsourcing sites/tools you have used for any language learning" In answer to Where do you use crowdsourcing sites, tools, and games to learn languages? participants from all countries mostly chose "Outside class" (POL=9 4.8%>B&H=84.1%>TUR=81.4%>MAC=73.2%). A much smaller number stated that they also use them in class (30.3%) or other places (6.1%). To the question Why have you used crowdsourcing websites, tools, and games? participants mostly responded with "for having fun while learning the language" (MAC=92.7%>POL =84.5%>TUR=79.1%>B&H=68.1%), "as a class activity" (38.4%), and/or "as a class assignment" (19.4%).
Finally, we asked our participants to indicate the FL they had learned while using crowdsourcing sites, tools, and games. They listed English, German, Spanish, French, Italian, and Turkish ( Figure 3). These results are almost parallel to the findings of recent studies (Luca, 2018), which reported that the most studied six FL in Europe are (1) English, (2) French, (3) German, (4) Spanish, (5) Russian, and (6) Italian. The only exception in our study was Turkish, which replaced Russian on the list. One reason for the observed difference might be the countries where the data were collected. Because of the Ottoman Empire, there are historical ties between B&H, MAC, and TUR. In B&H, Turkish is one of the FL taught in schools and universities (Ulutaş, 2018), while in Macedonia, there is a Turkish minority for whom Turkish is their heritage/mother language (Jašar-Nasteva, 2001). So, in both of these countries, in line with modern trends, LL use crowdsourcing tools to learn Turkish.

Conclusions
The results from our four culture/language diverse countries show that there are more similarities than differences in the ways in which LL in TUR, POL, MAC, and B&H perceive and employ online crowdsourcing resources: • the majority of respondents conceptualise crowdsourcing in the same way (Definition 2); • Wikipedia, despite being the oldest, is the most used crowdsourcing site (N=158; 74.9%); • contrary to the common perception of the pervasiveness of gaming among learners, 80.6% of our respondents did not use games to learn FL; • language learning is mainly done using smartphones and laptops, with minimal usage of other devices; • 63.5% of the participants use crowdsourcing websites/tools as an outsideof-class activity, and 79.6% use them to have fun while learning FL; • crowdsourcing sites/tools are mainly used to learn English (84.4%) as an FL.
Hopefully, the findings of this study can serve as guides for material developers/ experts/teacher trainers who strive to achieve crowdsourcing platforms/tools that are valid and appealing across cultures.