Understanding postgraduate students’ perceptions of plagiarism: a case study of Vietnamese and local students in New Zealand

Despite increasing scholarly interest in tertiary student perceptions of plagiarism, very little is known about those held by postgraduate (PG) students, although differences between undergraduate (UG) and PG students relate to both their characteristics and the demands of their studies. Furthermore, there is a dearth of research within the context of international education, where managing plagiarism is seen as a major challenge. This paper reports on a recent online survey with 207 Vietnamese (n = 72) and local (n = 135) PG students at a New Zealand university regarding their perceptions of plagiarism. The findings showed significant differences both between and within the two groups. Perception variations arose from a range of influences and prior experiences, not just the culture in which the students were initially educated. Differences related to participants’ age, gender, academic levels, disciplines, and teaching experience. This study’s findings contribute knowledge about under-researched PG students and problematize prevalent stereotypes of international students regarding plagiarism. They generate implications for higher education institutions to accommodate international and domestic PG student needs in ways that respect their diversities in detail as individuals, not as members of a homogeneous group. Further qualitative research to explore PG student perceptions in greater depth is recommended.


Introduction
Plagiarism is widely considered an intentional moral transgression (Pecorari & Petric, 2014;Sutherland-Smith, 2008) or a violation of academic integrity principles (Ehrich et al., 2016;Pecorari, 2016). In fact, it is a complex and multifaceted issue not only because of how it is defined but also what it covers and how it happens. While some researchers focus on textual features and define plagiarism as an issue of citation and acknowledgement (Hayes & Introna, 2005), others conceptualize plagiarism as responding to academia's ethical obligation to support international PG students' development as scholars. Effective support of these learners is an ethical matter because not only do Western universities typically charge higher prices for international students to undertake coursework, but also international (like domestic) PG students take on key roles after completion of their higher qualifications. This is illustrated by the Vietnamese government's provision of scholarships for early career Vietnamese academics. Their enhanced knowledge and skills gained through PG study are especially significant for the ongoing development of the higher education system in their home country.
While there is a growing breadth of research that explores student perception of plagiarism, the PG population has not been a focus (Du, 2020;Ramzan et al., 2012), particularly in international education contexts (Green et al., 2006;Orim et al., 2013). While there are very few investigations of plagiarism perceptions held by Vietnamese (Tran, 2012) and New Zealand students (Adam et al., 2016;Marshall & Garry, 2006), they focus mainly on UG students. Seeking to address this gap, this study examined the reality of PG students' awareness of plagiarism in a New Zealand university, including two groups: participants who obtained their previous tertiary education in Vietnam (Vietnamese students) or New Zealand (local students). The participants came from either individualist (local students) or collectivist (Vietnamese students) cultures and they speak English either as a first language (local students) or a foreign language (Vietnamese students). The study focuses not only on the PG students' perceived seriousness of plagiarism but also their understanding of related aspects contributing to intentional and unintentional plagiarism. We define plagiarism as "presenting someone else's words and/or ideas as one's own without proper attribution, either advertently or inadvertently". PG student judgments of plagiarism, including why it is wrong and its impacts, were the focus of the study. The research contributes insights into differences in attitudes and understanding of plagiarism of PG students from different educational, cultural and linguistic backgrounds. It allows deeper understanding of perceptions of plagiarism held by domestic and international PG students, which can inform development of interventions to support and maximize learning experiences for students with diverse backgrounds.

Plagiarism and international students in higher education
An increasing number of students study outside their home country, over 5.3 million students globally in 2019, many of whom are PG students (Institute of International Education, 2019). In the context of internationalization of higher education, managing plagiarism is perceived as a significant challenge facing Western higher education institutions (e.g., Bretag, 2013;Heitman & Litewka, 2011). Writing and referencing are challenging for international students because they transfer to a new academic context (Adhikari, 2018;Leask, 2006), and for many of them, English is a second or foreign language (Franken, 2013;Marshall & Garry, 2006;Pecorari, 2016). The literature reflects an ongoing examination of whether international students in general plagiarize more than domestic students and to what extent their understanding of conventions and expectations of academic writing differ. However, scholarship that illuminates knowledge Tran et al. International Journal for Educational Integrity (2022) 18:3 Page 3 of 21 related to PG students is very limited. Apart from this gap in the literature, an ongoing debate in the field relates to links between culture and plagiarism. Culture is a recurring topic within debate about the nature of links between culture and plagiarism by international students. Zobel and Hamilton (2002) found an association between rote-learning habits and UG students' perceived seriousness of plagiarism and Sowden (2005) argued that practices of multilingual students (e.g., communal ownership of knowledge, copying and reproducing practices) were in conflict with Western values related to plagiarism. However, Phan (2006) disputed an association between cultures and plagiarism attitudes, emphasizing that plagiarism was unacceptable in Vietnamese culture and Vietnamese students' memorization practices were unrelated to plagiarism. Liu (2005) reviewed books showing that plagiarism was prohibited in China, highlighting how stereotypes of non-Western cultures have led to inaccurate ideas about international students. Culture, while clearly important with regard to student preferences and experiences, cannot be regarded as a simple explanation for plagiarism behaviour, and its use raises questions about implicit bias in the treatment of students from different cultures.
Various determinants of plagiarism, other than simply cultural ones, have been identified. Some scholars suggest that language proficiency and experiences might substantially affect student practices, both as writers and in avoiding plagiarism (Marshall & Garry, 2006;Tran, 2012). Tasks requiring engaging with source materials can be doubly difficult for students with limited language resources (Schmitt, 2005). NESB students might engage in patch-writing when producing academic work (Pecorari, 2016). Educational backgroundeducational approaches, writing practices and assessment methodsmay impact students' knowledge, perceptions and experience of plagiarism (Adhikari, 2018;Marshall & Garry, 2006). These studies suggest that substantive issues in understanding plagiarism are experienced by all groups of students and processes that are responsive to specific groups (e.g., students from particular cultures or levels of study) would be more effective than initiatives aimed at the entire student population as if it was homogeneous.
Awareness of the limitations of cultural explanations of student plagiarism has seen attention shift to an alternative approach focused on understanding the issue beyond purely cultural influences. Scholars highlight the need for a holistic stance towards plagiarism by international students. Leask (2006) asserted that stereotyping Asian students as more susceptible to plagiarism hinders effective responses to the issue. Universities need to take into consideration the diversity of student backgrounds and employ rehabilitative as opposed to harsh punitive approaches to plagiarism management (Adam et al., 2016;Green et al., 2006;Marshall & Garry, 2005). Some suggest making plagiarism rules explicit to students, increasing their awareness of academic integrity and providing them with academic skills, rather than policing and punishing (Adhikari, 2018;Marshall & Garry, 2005;Gullifer & Tyson, 2010).
These remarks highlight the contributions of this study which attempted to explore how PG students from diverse educational, cultural and linguistic backgrounds differed in their understanding of plagiarism and factors associated with variations in their perceptions.

Domestic and international students' perceptions of plagiarism
A growing body of empirical research has examined domestic and international students' perceptions of plagiarism, revealing differences in perceptions between these Tran et al. International Journal for Educational Integrity (2022) 18:3 Page 4 of 21 two student cohorts, but most of these relate to UG students. This scholarship provides a body of knowledge that is a point of consideration for the examination of the PG student experience. Many international UG students possessed superficial understanding of plagiarism and referencing conventions, as demonstrated by Stappenbelt's (2012) study of firstyear international UG students in Australia and Bamford and Sergiou's (2005) research in the UK. Local UG students in Australia (Gullifer & Tyson, 2014) and the USA (Shi, 2004), in contrast, had better awareness of plagiarism and academic writing conventions. In a cross-cultural comparison study, Rinnert and Kobayashi (2005) revealed that American UG students were more likely than their Japanese counterparts to view plagiarism as unethical and showed greater awareness of the importance of acknowledging sources. Most Asian international students in a UK study did not view copying from textbooks as cheating (Bamford & Sergiou, 2005) and many international UG students in an Australian university considered that not plagiarizing was primarily to avoid penalties (Sutherland-Smith, 2008). While most Australian UG students described plagiarism as "dishonest", most international UG students considered plagiarists as inexperienced and just a few of them viewed plagiarism as dishonest (Stappenbelt, 2012). These consistent findings of different perceptions held by domestic and international students affirm the value of research that takes account of students' backgrounds, which would also be beneficial to develop understanding of PG students' experience.
International UG students viewed various types of plagiarism less severely than locals. Marshall and Garry (2006) found that NESB undergraduate students regarded most forms of plagiarism less seriously than their English-speaking background peers. Stappenbelt (2012) reported that most Australian first-year engineering students were aware of the seriousness of plagiarism and considered it an unacceptable practice whereas their international counterparts viewed plagiarism as a less severe offense. In a study of 2500 UG and PG students from multiple UK and Australian universities, Sutton et al. (2014) found that students previously educated in China and Southeast Asia viewed poorly referencing less seriously than those educated in Europe or Australia. Ehrich et al. (2016) revealed that more Chinese than Australian UG students accepted excuses for plagiarism and Australian students viewed plagiarism more severely than Chinese students.
Although differences are apparent between perceptions of plagiarism held by international and domestic UG students, we argue that it is inappropriate to make assumptions that findings for UG students are simply transferable to the experiences of PG students. PG students have different experiences and perspectives, including knowledge of plagiarism and academic integrity (Rinnert & Kobayashi, 2005;Sutton et al., 2014), and different career trajectories (Artess & Hooley, 2017). The danger of treating UG and PG students as homogenous is supported by studies which highlight that student perceptions of plagiarism change during their learning process, indicating a relationship between academic exposure and improvements in understanding of plagiarism and academic integrity. For instance, definitions of plagiarism of international students with longer engagement in Australian educational environments were closer to those of domestic students (Song-Turner, 2008). Similarly, experienced students were more conscious of textual identities than those with less writing experiences (Abasi et al., 2006) and students in the later years of their degree showed better understanding of Tran et al. International Journal for Educational Integrity (2022) 18:3 plagiarism than those transitioning to tertiary study (Hu & Lei, 2015;Stappenbelt, 2012). Therefore, level of study is a factor that affects student perceptions and understanding of plagiarism, highlighting the relevance of further scholarship to address the knowledge gap related to PG students.

Factors associated with students' perceptions of plagiarism
Educational background is a major influence on student perceptions. A large and growing body of literature has challenged deficit models of education regarding plagiarism by international students and identified alternative explanations, but this field too has neglected PG students thus far. Much of the literature reported that UG students' ability to identify plagiarism varied across educational contexts (Green et al., 2006;Marshall & Garry, 2006;Stappenbelt, 2012). Domestic and international students may have received dissimilar prior plagiarism education which may lead to different perceptions (Leonard et al., 2015;Stappenbelt, 2012). Assessment methods focusing on textbook content discouraged UG students' critical thinking and expression of ideas (Hayes & Introna, 2005) and examoriented and written assessments contributed to their plagiarism (Bennett, 2005;Song-Turner, 2008). Limited exposure to coursework might influence students' views of academic integrity and their use of source materials (Hayes & Introna, 2005;Nguyen & Buckingham, 2019). These studies suggest the role of learning design as a factor affecting student perceptions.
Explicit plagiarism training makes a difference. A training session on referencing improved Chinese students' knowledge of referencing and plagiarism (Du, 2020) and a 13-week course on plagiarism-related issues enhanced international students' academic writing skills and understanding of plagiarism (Tran, 2012). Perkins and Roe (2020) revealed the effectiveness of an academic English master class on Vietnamese students' understanding of academic conventions.
Raising student awareness of the university plagiarism policies can enhance their understanding of the issue. Students who received information about plagiarism and its avoidance perceived plagiarism more seriously and were less likely to plagiarize than those who did not (Brown & Howell, 2001). Male students, either UG or PG, were more likely to read policies (Gullifer & Tyson, 2014), and possessed slightly more tolerant (Bokosmaty et al., 2019) and positive attitudes toward plagiarism (Jereb et al., 2018) than their female counterparts. Hu and Lei (2015), however, reported no significant difference between male and female students.
Research suggests that policies need to reflect disciplinary diversity within the university, with regard to plagiarism. Students from different disciplines have been shown to hold different views on plagiarism. Business students, both UG and PG, considered plagiarism less seriously than those from other disciplines (Sutton et al., 2014). UG students majoring in English language and business studies were more likely than those in mechanical engineering and computer engineering to view slack attitudes as causes of plagiarism (Hu & Lei, 2015).
These studies point to the complexity of plagiarism and suggest that more work is needed to gain more holistic understanding of the issue. Findings from these studies call for a shift from stereotyping international UG students regarding plagiarism to realizing and acknowledging their diversities to better accommodate their needs, and we assert that this principle is equally valid for international PG students.

Postgraduate students' perceptions of plagiarism
Studies showed contradictory findings regarding PG students' understanding and perspectives of plagiarism; we note that investigations have mainly been conducted in non-Western contexts. Most Iranian graduate students considered plagiarism more as unintentional behaviour than a deliberate attempt to cheat (Babaii & Nejadghanbar, 2017;Rezanejad & Rezaei, 2013). Du (2020) found that most Chinese PG students possessed limited understanding of plagiarism before engaging in a training session about referencing and plagiarism. Limited understanding of plagiarism was held by PG students in Japan (Rinnert & Kobayashi, 2005) and Pakistan (Ramzan et al., 2012). Selemani et al. (2018), in contrast, reported that all PG students in a Malawian university considered plagiarism as a serious academic breach. There is much scope in the field for further knowledge development of PG students' understandings and perspectives of plagiarism beyond these settings.
Few studies have explored PG students' perceptions of plagiarism in international education contexts and these revealed inconsistent findings regarding understanding of plagiarism for both domestic and international students. Most Nigerian PG students were found to possess limited prior knowledge of plagiarism and came to realize the seriousness of plagiarism when commencing their PG programmes in a UK university (Orim et al., 2013). Leonard et al. (2015), in contrast, found that most PG students previously educated either outside or within the USA viewed plagiarism as a serious issue. Gullifer and Tyson (2014) reported good understanding of plagiarism among most domestic PG students in Australia and Green et al. (2006) found that they were better than international students in identifying plagiarism. In one study which investigated the impact of an intervention on PG student plagiarism, plagiarism by international master's students at an Australian university greatly reduced after the intervention, implying that in this case too, explicit training was useful (Duff et al., 2006).
In three studies that we found, similarities and differences regarding perceptions of UG and PG students were observed. Ryan et al. (2009) revealed that few UG and PG students in Australia knew what their university's policy of plagiarism covered although most were aware of the policy. They also reported inadequate understanding of plagiarism among most students at both UG and PG levels. In other studies, PG students viewed poor referencing more seriously (Sutton et al., 2014) and perceived proper citation as more crucial than did UG students (Rinnert & Kobayashi, 2005). These contradictory findings suggest that factors related to level of study need further exploration.
In summary, our review of literature found that the range of research settings and participants investigated thus far leave room for further knowledge development. While a number of studies focused on PG students' perceptions of plagiarism, most were conducted in non-Western contexts. Research that specifically targets the PG student population in international education settings is still scant (Green et al., 2006;Leonard et al., 2015;Orim et al., 2013). Also, most studies simply compared perceptions between domestic and international students without exploring other factors associated with their perceptions. There remains little data regarding international PG students' perceptions of plagiarism, how their perceptions differ from locals, and what leads to any differences.
Our review of the literature fueled our curiosity to learn how these findings and recommendations related to both domestic and international PG students in the New Zealand context. We had a specific interest in Vietnamese-educated PG students, because Author 1 is a Vietnamese academic who, like many of her peers, has gained a government scholarship for her PhD study in New Zealand, as part of the nation's strategy to strengthen Vietnamese higher education institutions. This work was framed by a conception of the plagiarism space that recognizes that the diverse experiences of university education and life in different countries and cultures must have a significant impact on individual student's knowledge and capabilities in using scholarly information. This diversity, we posit, must be apparent in a heterogenous conception of plagiarism in student populations that transcends the simple domestic/international framing of much of the plagiarism literature presented above.

Research questions
This study explored Vietnamese and New Zealand postgraduate students' perceptions of plagiarism focusing on their educational experiences. The main research question was: How do Vietnamese and New Zealand PG students perceive plagiarism?
Two sub-questions framed this investigation: a. How do students' attitudes and understanding of plagiarism differ between these groups? b. How do students' attitudes and understanding of plagiarism differ within these groups?
Ethical approval for this study was obtained from the Human Ethics Committee of the university prior to data collection (approval number 0000025145).

Research instrument
An original online questionnaire was developed and used to collect data. Pre-existing scales were not utilized because they were not relevant to the scope of the study. The initial questionnaire included (1) four demographic questions, (2) three questions about students' educational backgrounds, and (3) a Plagiarism Perception Scale with 39 Likert scale items constructed by undertaking a thorough literature review on students' perspectives of plagiarism.
In the development of this research instrument, several strategies were employed to ensure its validity and reliability. To establish its content validity, three experts in a New Zealand university evaluated the extent to which each item was relevant to the research questions. All the three experts have experience in teaching and supervising international and domestic postgraduate students. One expert specializes in education, one in quantitative research and the other in plagiarism and academic integrity. Following expert review and revision, ten international PG students at the same university who were ineligible for study participation were invited to pretest the second version of the questionnaire to determine if they could understand and answer the questions easily.
After this process, version three, excluding demographic questions, was piloted with 32 international volunteer PG students to explore how well they understood the questionnaire items and identify problematic items. Data were analyzed using both descriptive statistics and Spearman's Correlation. Descriptive statistics were performed to explore the distributions of responses for each survey item. Items that mostly received extreme response options were closely examined as to whether they represented misunderstanding or exaggeration from the respondents. Spearman's Correlation was used to explore correlations within the data set and to examine if the students consistently answered the survey. Results demonstrated that most items, which were anticipated to be correlated, had monotonic relationships with each other, indicating that they were consistently answered. However, some examples here, which we expected to be related, appeared to have no significant correlations, suggesting that some concepts might not be fully understood by respondents. Three items, 2-Plagiarism happens when students are learning to write, 8-Plagiarism is common in the work of novice writers, and 26-Students' poor understanding of how to write academically causes plagiarism, described the relationship between writing ability and plagiarism. While there was a positive correlation between items 2 and 8 (rs = .495, p = .004); 8 and 26 (rs = .370, p = .037), the correlation coefficient value of .229 suggested no correlation between items 2 and 26. These items were rewritten following the experts' suggestions.
Drawing on the analysis, 15 items were reworded to make them more concise. Three items were removed and one was added to the survey. Modifications of sentence structure were also made. The pilot data analysis resulted in the selection of 37 items for the actual survey. The final questionnaire with a six-point Likert Scale in which 1 = Strongly agree, 2 = Agree, 3 = Slightly agree, 4 = Slightly disagree, 5 = Disagree, and 6 = Strongly disagree was distributed using Qualtrics survey software.
A copy of the questionnaire is provided in the Appendix. The results of factor analysis will be available on request.

Data analysis
Before the main analyses, exploratory factor analysis was performed to determine the underlying constructs of the data set. Factor analysis with maximum likelihood estimation and oblique rotation was performed with all 37 items. The KMO statistics was 0.733 and Bartlett's Test of Sphericity was significant (χ2 = 1572.60, df = 406, p < .001). After multiple analyses, 29 items were retained, loading into five factors and accounting for 44% of the variance. Factor 1-Perceived seriousness of plagiarism (α = .69) accounted for 14.2% of the variance and included eight items representing different degrees of plagiarism. Factor 2-Personal factors (α = .70) consisted of four items describing personal factors for student plagiarism, accounting for 11.3% of the variance. Factor 3-Academic factors (α = .70) accounted for 7.9% of the variance and comprised six items. Factor 4-Assessment factors (α = .59), including three items describing assessment factors, accounted for 6% of the variance. Factor 5-Negative attitudes towards plagiarism (α = .68) included six items reflecting negative attitudes towards plagiarism and accounted for 4.6% of the variance. The Cronbach's alpha value for the Plagiarism Perception Scale was .73, well within the range expected for a reliable instrument.
Independent samples T-tests were used to determine if there were statistically significant differences in perceptions between students who were (1) Vietnamese and New Zealanders, (2) male and female, (3) PhD and Master's students, and (4) students with and without teaching experience. The Levene's tests for homogeneity of variances indicated that the variances were equal across two groups in all five subscales, (p values > .05). Therefore, t-test results associated with the criterion 'Equal variances assumed' were reported. ANOVA was employed to compare means among (1) disciplines and (2) age groups. The number of respondents for each demographic and educational question differed, so the population (n) in each of the analyses varied and represented the total number of respondents answering individual questions.

Research sub-question 1
An independent-samples t-test was conducted to compare Vietnamese and New Zealand PG students' perceptions of plagiarism, shown in Table 1.
In subscale 1-Perceived seriousness of plagiarism, results indicated a significant difference, (t(205) = 3.982, p < .05), with New Zealand PG students (M = .4666, SD = 1.62739) viewed plagiarism more seriously than their Vietnamese counterparts (M = −.4646, than Vietnamese students (19.4%) strongly disagreed with the concept that selfplagiarism was not a serious offence. While self-plagiarism was regarded as the least serious type of plagiarism by New Zealand students, the Vietnamese students viewed patch-writing as the least severe. Both groups viewed direct copying as the most severe form of plagiarism. There was a significant difference (t(205) = 2.777, p < .05) in subscale 3-Academic factors, suggesting that Vietnamese students (M = −.3592, SD = 1.50932) were more likely than New Zealand students (M = .2425, SD = 1.47166) to perceive that plagiarism resulted from the lack of knowledge about plagiarism and limited academic skills. Distributions of students' responses for individual items showed that while 83.4% of Vietnamese students agreed that students plagiarized because they could not tell what was or was not plagiarism, 74.4% of New Zealand students agreed with that idea. 86.4% of New Zealand students agreed that students' poor understanding of citing and referencing led to plagiarism whereas 91.7% of Vietnamese students agreed with the statement.
Another significant difference was found in subscale 4-Assessment factors, (t(205) = 3.982, p < .05), with Vietnamese students (M = −.5222, SD = 2.40846) agreed more strongly than New Zealand students (M = .2979, SD = 2.11261) that plagiarism resulted from assessment factors. There was no significant difference between the two groups in subscales 2-Personal factors and 5-Negative attitudes towards plagiarism, suggesting that both groups considered that plagiarism resulted from personal factors and they held negative attitudes towards plagiarism. Distributions of students' responses for individual items indicated that most respondents agreed with the idea that plagiarism damaged the meaning of a degree (over 98%), and that it was deceitful (95.1%) and stealing (91.3%). Over 78% of them agreed that plagiarizing students learnt less and nearly 83% agreed that plagiarism worsened teacher-student relationships. However, 46.9% of them agreed that a good person did not plagiarize.

Research sub-question 2
Significant differences in the PG students' perceptions in relation to their demographics were apparent. The means of responses from the gender groups on the five subscales are compared in Table 2. With only three non-binary students and one genderqueer student, the comparison was conducted with males and females, who accounted for most of the sample. Table 2 shows a significant difference between two groups, (t(201) = 0.21, p < .05) in subscale 5-Negative attitudes towards plagiarism, suggesting that female PG students (M = .2384, SD = 1.58) viewed plagiarism more negatively than males (M = −.4206, SD = 1.48). There were no significant differences between males and females in subscales 1-Perceived seriousness of plagiarism, 2-Personal factors, 3-Academic factors, and 4-Assessment factors, suggesting that male and female PG students were not different in their perceived Tran et al. International Journal for Educational Integrity (2022)  seriousness of plagiarism and their levels of agreement on different determinants of student plagiarism. Table 3 presents the means of responses from the four age groups: younger than 24, 25 to 34, 35 to 44, and older than 44.
In Table 3, differences in PG student perceptions of plagiarism between age groups were not significant (p values > .06) for subscales 2-Personal factors, 3-Academic factors, 4-Assessment factors, and 5-Negative attitudes towards plagiarism. These results suggest that PG students across age groups agreed on factors contributing to plagiarism and viewed it negatively. For subscale 1-Perceived seriousness of plagiarism, the means were statistically different (F(3) = 3.092, p < 0.05). Figure 1 illustrates how the mean varied between different groups.  The mean plot shows a clear difference between the views of PG students who were less than 24 years old (M = −.1824, SD = 1.45322) and those who were older than 44 (M = .8493, SD = 1.74546), with the latter group viewing it more seriously.
The means of responses from Masters' and PhD students were compared in Table 4. Table 4 shows a statistically significant difference in Subscale 1-Perceived seriousness of plagiarism, (t(193) = − 2.276, p < .05), with PhD students (M = .4176, SD = 1.73431) viewing plagiarism more seriously than Master's students (M = −.1243, SD = 1.58200). Another significant difference (t(193) = − 2.554, p < .05), was found in subscale 4-Assessment factors, with master's students (M = −.3876, SD = 2.17117) being more likely than PhD students (M = .4326, SD = 2.30554) to consider features of assessment as causes of plagiarism. There were no statistically significant differences between the two groups for subscale 5, suggesting that both groups held negative attitudes towards plagiarism. Differences in PhD and Master's student perceptions of plagiarism were not significant in subscales 2-Personal factors and 3-Academic factors, representing that both groups agreed that plagiarism resulted from personal factors and limited academic knowledge and skills.   Table 5 shows a comparison of the means of responses from students from the four faculties comprising the most respondents: Humanities and Social Sciences, Business, Science, and Education.
In Table 5, the one-way ANOVA comparisons were not significant (p values > .06) for subscales 1-Perceived seriousness of plagiarism, 2-Personal factors, 3-Academic factors, and 4-Assessment factors. In subscale 5-Negative attitudes towards plagiarism, the means of the four discipline groups were statistically different (F(3) = 4.184, p < 0.01). Therefore, in Fig. 2 below, the four means were plotted with standard errors to observe the differences.  The mean plot indicated a significant difference between Education students (M = .7807, SD = 1.39881) and Science students (M =. -4937, SD = 1.57467), suggesting Science students possessed more negative attitudes towards plagiarism than Education students.
The means of responses from PG students who had or had not engaged in teaching jobs are compared on five subscales in Table 6. Table 6 shows a statistically significant difference (t(205) = 2.095, p < .05) between students with teaching experience (M = −.0963, SD = 1.58474) and those without (M = .4001, SD = 1.20840) in Subscale 3, indicating that PG students with teaching experience were more likely than those without teaching experiences to perceive academic factors as causes of student plagiarism. There were no statistically significant differences between the two groups in subscales 1-Perceived seriousness of plagiarism, 2-Personal factors, 4-Assessment factors, and 5-Negative attitudes towards plagiarism.

Discussion
This study found significant areas of agreement between the Vietnamese and local PG students in New Zealand. Most Vietnamese and New Zealand PG participants held substantial understanding of plagiarism, both in terms of its meaning and complexity. They could distinguish between plagiarism and acceptable practices. The students held negative attitudes towards plagiarism, considering it as deceitful and as stealing. The findings are different from previous studies showing that international students did not view plagiarism as an ethical issue (Stappenbelt, 2012;Sutherland-Smith, 2008). While agreeing that students deliberately plagiarized because of academic pressure and unearned benefits, many understood reasons that might lead to unintentional plagiarism such as limited understanding of plagiarism, academic writing, and referencing conventions. The findings are aligned with published research showing PG students' conceptual understanding of plagiarism (Selemani et al., 2018) and corroborate previous findings indicating that most PG students considered plagiarism as academically serious (Gullifer & Tyson, 2010;Selemani et al., 2018). In contrast with findings from UG contexts (Du, 2020;Ehrich et al., 2016), this sample of PG students did not show confusion or superficial understanding of plagiarism. While existing literature is limited to UG students or PG students in non-Western contexts, unique findings from this study emphasize the importance of studying international and domestic PG students' experiences in an international setting and not assuming that findings from UG contexts are relevant to other cohorts in another setting. Vietnamese and New Zealand PG students' responses to three out of five subscales were significantly different, confirming a link between educational backgrounds and plagiarism perceptions identified in other student populations (Bennett, 2005;Marshall & Garry, 2006). New Zealand PG students viewed plagiarism more severely than their Vietnamese peers, supporting the idea that local students viewed several types of plagiarism more seriously than international students (Marshall & Garry, 2006;Sutton et al., 2014). Vietnamese agreed more strongly than New Zealand students with academic and assessment factors for plagiarism, which may be due to their dissimilar educational approaches, academic writing experiences, and assessment practices (Marshall & Garry, 2006;Song-Turner, 2008). Some differences could arise from their assessment tasks and how students have approached these, at UG and/or PG level. The Vietnamese students viewed patch-writing as the least severe likely because patch-writing is often used to learn English writing (Pecorari, 2016). Also, while New Zealand higher education focuses on critical reading and thinking, memorization is a common learning strategy in Vietnam (Phan, 2006). As we discussed earlier, many international PG students may take on important societal roles when they return with higher qualifications to their home countries. The findings suggest the importance of university initiatives. As well improving PG students' awareness when they are studying overseas, these may inform future plagiarism education initiatives to which higher education graduates may contribute in their home countries, thus supporting the realization of national goals which underpin government expenditure for international PG study.
In this study, Master's and PhD students differed in their perceptions about the link between features of assessment and plagiarism which may be due to assessment practices that students have experienced. In New Zealand, while master's programmes could be by either coursework or thesis, PhD students are required to produce theses. PhD students viewed plagiarism more seriously than Master's students, consistent with previous evidence of a relationship between academic levels and student understanding (Rinnert & Kobayashi, 2005;Sutton et al., 2014). Another possible factor is students' length of exposure to academic writing and university initiatives aiming at addressing plagiarism (Du, 2020;Perkins & Roe, 2020). While PhD students are committed to their research for three to four years, Master's students complete their studies within a shorter timeframe. Differences between Master's and PhD students' perceptions are one example of several differences that were apparent within the two groups of Vietnamese and New Zealand PG students, suggesting the complexity of student perceptions and how they arise.
Other connections were found between demographics and PG student perceptions of plagiarism. Females held more negative attitudes towards plagiarism than males, supporting the association between gender and plagiarism perceptions in the literature (e.g., Bokosmaty et al., 2019;Jereb et al., 2018). The finding, however, counters previous research which found no significant differences in perceptions of plagiarism held by male and female students (Hu & Lei, 2015). Older PG students perceived plagiarism more seriously than those who were younger, which is likely a consequence of writing and academic experiences (Abasi et al., 2006;Hu & Lei, 2015). Students from the faculty of science possessed more negative attitudes than education students, confirming that student perceptions are associated with their disciplinary norms, referencing practices and conceptualization of plagiarism (Rinnert & Kobayashi, 2005;Sutton et al., 2014). The finding can be explained drawing on the fact that teaching experiences exposed participants to student plagiarism which gave them a greater insight into reasons for both intentional and unintentional plagiarism. PG students with earlier employment as a teacher or tutor agreed more strongly than those without teaching experience on academic factors for plagiarism, confirming an association between professional experiences and plagiarism perceptions (Leonard et al., 2015). This finding is interesting in the context of the debate between those treating plagiarism primarily as misconduct or cheating, and those that see it as an educational opportunity. Those adopting a punitive perspective would expect emerging teachers to be intolerant of plagiarism, while an educational view would position this as representing an aspect of the zone of proximal development at play as emerging teachers respond to the complexities of academic information use. Unique findings of intra-group differences found within this study's sample of Vietnamese and New Zealand PG students studying in New Zealand problematize approaches that treat students as homogenous, and highlight the value of additional research to further explore perceptions of PG students, both domestic and international.

Opportunities for further research
This study identified differences in domestic and international PG student perceptions, but not where those differences came from. Additional qualitative research could develop rich knowledge of how student perceptions develop. The survey examined student perceptions at one point in time; however, longitudinal research with PG students in university programs could show how their understanding evolves and the dynamic nature of influences, including university strategies. Research with larger sample sizes is needed to explore transferability of findings to other academic contexts.
Although the questionnaire has been through various validation processes (expert reviewing, pretesting, piloting and exploratory factor analysis), further validation by means of confirmatory factor analysis might be added before using the tool. While the overall alpha for the Plagiarism Perception Scale with 29 items was .73, well within the expected range for a reliable instrument, the alpha value for Subscale 4 with three items was relatively low (.59) and this may be able to be improved in the future with additional items. Future researchers may also invite international experts to review the questionnaire before using it.

Conclusion
The research identifies differences in PG student perceptions that arose from multiple factors rather than simply the system in which they were previously educated. The findings confirm that culture cannot be regarded as a simple explanation for plagiarism perceptions or behaviours, challenging implicit bias in the treatment of students from different cultures.
The contribution of the study relates to knowledge both for PG students and in regard to the Vietnamese international student experiences, which have not previously been studied. Findings of both inter-and intra-group differences in perceptions of Tran et al. International Journal for Educational Integrity (2022) 18:3 plagiarism held by Vietnamese and local PG students at a New Zealand university highlight that students are diverse, and their needs and expectations are dynamic with various academic contexts embedded in different cultures. Therefore, development of deeper understanding of different student contexts may avoid reliance on misleading assumptions about their perceptions and help to develop more comprehensive and relevant support resources. The findings support moving beyond merely policing policies or focusing on policing adherence to such policies, towards taking a more educative approach. The study highlights the potential of applications in universities of a programme of differentiated learning opportunities about plagiarism because PG students and their learning needs are not homogenous. Enactment of such an approach would mean that academics acknowledge students' differing starting points and provide them with needs-based support to enhance and maximize their learning experience.
With an increasing number of international students globally, especially PG students, plagiarism has become an increasingly complicated issue. Our findings show that the diverse prior experiences of these students are powerful influences on their perceptions and priorities. There are clearly differences arising from factors such as English language skills, such as the Vietnamese students more permissive attitude to patchwriting, but also strong influences based on level of study as were found between Masters and PhD students, and on the experience of formal teaching roles. This latter difference is important to consider as a factor when engaging with doctoral students in the international space as many countries use international doctoral scholarships as professional development and capability building for staff teaching in national educational systems. There is also the opportunity for universities to use sessional teaching experiences during doctoral study to expose PG students to the challenges faced by others in using scholarly information with integrity.
By revealing intra-group differences in PG student perceptions, the study highlights the complexity of PG student experiences that might impact their perceptions, problematizing approaches that simply compare international and local students, and challenging discourses about plagiarism which stereotype international PG students. To support the effective and ethical delivery of PG studies in the global education model, we call for further work that deeply illuminates PG student perspectives and explores points of connection and difference across domestic and international PG student cohorts.