Evaluative That in Nursing Research Abstracts: A Comparative Study of International and Japanese Journals ()
1. Introduction
Recent advancements in artificial intelligence, combined with the borderless nature of information, have expanded opportunities for non-native English-speaking researchers to submit articles in English to international journals (Wang, Zhang, Jin, Xu, & Wang, 2024). However, academic disciplines adhere to distinct writing conventions unique to their respective fields (Skorczynska & Carrió-Pastor, 2021). Moreover, as Lu, Casal, Liu, Kisselev and Yoon (2021) argue, these conventions are not necessarily consistent within a given discipline, highlighting the need for further research even in the same academic field.
For example, different journals may apply varying standards, and these conventions are not always explicitly outlined in journal guidelines. In such cases, authors must infer unstated norms by analyzing multiple published articles. Native English speakers with high proficiency can better access and comprehend many articles and, consequently, are more likely to recognize these implicit standards. Hence, studies comparing English papers by non-native and native speakers have found the former to be less proficient in meeting such criteria (Al-Zubeiry & Al-Baha, 2019).
To address these challenges, research has increasingly focused on identifying the distinctive features of academic writing across disciplines, such as rhetorical movement (Peacock, 2002), impersonality (Hyland, 2001), critical appraisal (Côté & Turgeon, 2005), and stance (Biber, 2006).
Biber (2006) identifies various linguistic elements that convey stance, reflecting the author’s attitude, certainty, or evaluation of a proposition. A key element in this regard is using modal and semi-modal verbs, such as “might,” “must,” “can,” and “should,” which indicate degrees of certainty, necessity, or possibility. Stance adverbials, including “certainly,” “unfortunately,” and “frankly,” mark epistemic judgment, personal attitude, or communicative style. Additionally, extraposed clauses, such as “It is clear that he succeeded…,” provide an impersonal yet structured way of expressing stance. Lastly, stance complement clauses, formed with controlling verbs, adjectives, or nouns, explicitly frame an evaluation, as in “I believe that this is true.” The evaluative that clause, a specific stance complement structure, has received considerable research attention across academic disciplines (Hyland & Jiang, 2018; Alonso-Almeida, Alvarez-Gil, Moskowich, Lareo, & Camina, 2021).
The evaluative that clause is a grammatical structure used to express the author’s evaluation or position in academic writing. This structure comprises a main clause containing an evaluative predicate, such as the verbs “demonstrate”, “show”, or “find”, or an evaluative noun or adjective, followed by a subordinated that-clause as the object of evaluation. According to Hyland and Tse (2005), this structure serves two key functions: it explicitly conveys the author’s evaluation and guides the reader’s interpretation. Additionally, the evaluative clause clarifies the author’s stance and shifts responsibility for evaluation (Kim & Crosthwaite, 2019).
The frequency of this structure varies across academic disciplines and depends on the author’s proficiency in academic writing. For instance, in business and medical fields, authors frequently use this structure to evaluate both their own findings and existing research (Kim & Crosthwaite, 2019). In applied linguistics research, it is commonly employed to interpret results and compare them with other studies, particularly in discussion sections (Phongjit & Gampper, 2023). Studies indicate that second-language (L2) authors often lean toward either excessively definitive or overly cautious evaluations, in contrast to the more balanced and restrained approach characteristic of native speakers (Hyland & Tse, 2005).
This discrepancy may be, at least in part, from the emphasis on objectivity in academic writing for English as a Foreign Language (EFL) authors and novice scholars. Many universities provide explicit guidelines on maintaining objectivity in academic writing (University of Western Australia, n.d.; University of Melbourne, n.d.; Charles University, n.d.). The University of Western Australia, for example, highlights that objectivity enhances credibility and persuasiveness in academic writing. According to its guidelines, objectivity is achieved by avoiding personal judgments (e.g., “I think,” “I believe”), emotive language (e.g., “ridiculous,” “awesome”), and superlative adjectives (e.g., “the best,” “the worst”). Instead, authors must use neutral expressions, modal verbs, and precise evaluative terms. Providing specific evidence and maintaining a balanced tone are also key strategies for achieving objectivity. Similarly, Sakai, Shibata and Inoue (2018) observed that academic writing guidebooks in Japan emphasize the same principles.
Thus, it is hypothesized that EFL and novice academic writers can improve their effectiveness by developing proficiency in using evaluative clause structures appropriately. Moreover, understanding conventional usage patterns within their discipline can help them align their writing style with the expectations of their academic community (Phongjit & Gampper, 2023). This study is the first to examine differences in evaluative that clause usage between researchers publishing in EFL journals and internationally skilled researchers by analyzing abstracts from Japanese nursing journals and internationally recognized academic publications, thereby addressing a gap in nursing research on this topic.
2. Literature Review and Hypotheses
This section formulates hypotheses regarding potential disparities in evaluative that clause usage in nursing research abstracts, drawing on prior studies. Specifically, we examine whether authors publishing in Japanese nursing journals exhibit distinct patterns compared to researchers publishing in top-tier international academic journals. Additionally, we seek to identify the nature and extent of these differences, if they exist.
The most comprehensive and widely recognized analytical framework for evaluative that clauses is the model developed by Hyland and Jiang (2018), which builds upon the foundational work of Hyland and Tse (2005) (see Table 1). This framework is structured around four key analytical dimensions: evaluated entity, evaluative stance, evaluative source, and evaluative expression, providing a systematic approach to analyzing evaluative functions within academic discourse. The evaluated entity refers to what is being assessed, such as claims, research findings, or established knowledge. The evaluative stance captures the author’s attitude, classified as epistemic (certainty or doubt) or attitudinal (affective or obligation). The evaluative source identifies whether the evaluation originates from the author, other scholars, abstract concepts, or concealed sources. The evaluative expression considers how evaluation is conveyed, including verbal (e.g., cognitive, research, or discourse acts) and nonverbal (e.g., nouns and adjectives) forms. This framework has been widely applied in subsequent research on evaluative that clauses to identify discipline-specific characteristics (Kim & Crosthwaite, 2019; Deng, Wang, & Gao, 2023).
In the following section, we explain the rationale for the hypotheses presented in Table 2 regarding evaluative that clauses found in abstracts of internationally recognized nursing journals and Japanese nursing journals. The analysis is based on the four perspectives outlined in the framework proposed by Hyland and Jiang (2018).
Table 1. Classification of evaluative that clauses (modified from Hyland & Tse, 2005, p 130; Hyland & Jiang, 2018, p 8).
Aspect |
Sub-categories |
Examples |
Evaluated Entity |
Interpretation of the author’s claim |
Our findings show that the museum visit can be seen as six distinct stages. |
Interpretation of previous studies |
One limitation of Fransen et al. (2014) was that participants were only asked to evaluate the best leader on their team. |
Interpretation of the author’s goals |
It is our hope that the framework can be applied to other contexts. |
Interpretation of methods, models, and theories |
It was found that the results of the model were in very good consistency with... |
Common or accepted knowledge |
It is believed that there is a direct relationship between poverty and crime. |
Evaluative Stance |
Attitudinal effect |
I hope that/It is important to note that |
Attitudinal obligation |
It must be recognized that… |
Epistemic: Certainty |
This shows that… |
Epistemic: Doubt |
Our findings suggest that… |
Neutral |
This means that… |
Evaluative Source |
Author |
We show that/I indicate that… |
Other humans |
Smith notes that… |
Abstract entity—inanimate source |
The findings indicate that… |
Concealed—source not identified |
It is well-known that/A general finding is that… |
Evaluative Expression |
Nonverbal noun predicate |
We make the assumption that… |
Nonverbal adjectival predicate |
It is possible that/It is well-known that… |
Verbal predicate: research acts in the real world |
This demonstrates that/The analysis indicates that/They found that… |
Verbal predicate: discourse acts |
We argue that/I propose that… |
Verbal predicate: cognitive acts |
They perceive that/We believe that… |
Table 2. Hypotheses regarding the four categories between high impact factor international journals and Japanese journals.
Category |
International Journals |
Japanese Journals |
Evaluated Entity |
HI1: Primarily evaluates its own findings, with minimal reference to prior research. |
HJ1: Even less engagement with prior studies, focusing on self-evaluation. |
Evaluative Stance |
HI2: Epistemic doubt predominates, with certainty less pronounced, and attitudinal markers rarely observed. |
HJ2: More epistemic certainty, less epistemic doubt, with frequent attitudinal obligation. |
Evaluative Source |
HI3: Abstract entities most used; concealed sources and authorial evaluation secondary. |
HJ3: Greater reliance on abstract entities, with frequent use of concealed sources. |
Evaluative Expression |
HI4: Comparatively higher frequency of discourse verbs over research verbs, with cognitive verbs and adjectives rarely found. |
HJ4: Frequent use of research verbs followed by discourse verbs and nouns, with adjectives and cognitive verbs scarce. |
2.1. Hypotheses on the Evaluated Entity in Nursing Journal Abstracts
Preliminary findings from medical studies suggest that evaluative that clauses in nursing journal abstracts primarily emphasize on the authors’ own findings rather than on evaluations of previous research, methodological frameworks, or commonly accepted knowledge (Kim & Crosthwaite, 2019). Given the empirical and practice-oriented nature of nursing research, authors must emphasize their own results to highlight their contributions to clinical practice and patient care (Melnyk, Fineout-Overholt, Stillwell, & Williamson, 2009). This aligns with prior research indicating that medical research articles predominantly prioritize findings over other forms of evaluation.
Furthermore, studies on evaluative entities in EFL academic writing suggest that authors in EFL contexts follow a similar pattern, prioritizing self-evaluation over engagement with previous research or theoretical frameworks. For example, Hasan and Alsout (2021) found that Libyan authors predominantly evaluated their own findings and claims, with significantly fewer instances of assessing previous studies. This trend suggests that academic publications by EFL scholars tend to engage less critically with external sources. This may be due to linguistic limitations, academic writing conventions in EFL contexts, or a preference for emphasizing their own contributions.
2.2. Hypotheses on Evaluative Stance in Nursing Journal Abstracts
Findings from medical research provide a basis for the hypothesis that epistemic stance, particularly expressions of doubt, most frequently manifest as evaluative stance in nursing journal abstracts (Kim & Crosthwaite, 2019). Epistemic certainty occurs to a lesser degree, whereas attitudinal affect, attitudinal obligation, and neutral stance appear only minimally. Research indicates that medical and nursing disciplines prioritize empirical evidence and require cautious interpretation of results (Shen & Tao, 2021). Consequently, epistemic doubt (e.g., “suggest,” “may indicate”) may be dominant, allowing authors to present findings tentatively while avoiding overgeneralization. While epistemic certainty (e.g., “demonstrates,” “confirms”) is present, it may be less frequent due to the nature of scientific research, which often necessitates hedging. A comparison with internationally recognized journal articles (Abbasi Montazeri, Jalilifar, & Hita, 2023) reveals that applied linguistics research articles also favor epistemic stances, with epistemic doubt exceeding epistemic certainty. This pattern suggests a broader academic convention of mitigating claims, possibly also applicable to nursing research. Furthermore, attitudinal affect (e.g., “it is important that”) and obligation (e.g., “it must be recognized that”) appear in international journal articles but remain secondary to epistemic stances. Given the emphasis on practical and clinical applications in nursing research, attitudinal stance remains pertinent, though its frequency is likely to be lower than that of epistemic stances (Melnyk, Fineout-Overholt, Stillwell, & Williamson, 2009). The neutral stance, exemplified by phrases such as “this means that,” which is rarely observed in medical and applied linguistics journals, may be even less prevalent in nursing abstracts.
Regarding authors published in Japanese nursing journals their use of the evaluative stance can diverge from that of researchers publishing in international high impact factor journals. Hasan and Alsout’s (2021) findings indicate that authors from EFL contexts frequently employ epistemic certainty while exhibiting minimal use of epistemic doubt. This may be from limited exposure to academic conventions that prioritize hedging. Researchers publishing in Japanese academic journals may, therefore, exhibit a stronger reliance on epistemic certainty (“proves that,” “confirms that”) than their native counterparts, leading to a more assertive style.
In sum, while nursing abstracts in top-tier international academic journals must prioritize epistemic doubt, authors published in Japanese nursing journals may diverge from this pattern by favoring epistemic certainty due to linguistic and cultural factors influencing their academic writing style.
2.3. Hypotheses on Evaluative Source in Nursing Journal Abstracts
Findings from medical journals (Kim & Crosthwaite, 2019) suggest that the evaluative sources in nursing abstracts will predominantly rely on abstract entities (e.g., “The findings suggest that…”) rather than authorial evaluation, as in “We argue that…”, given the emphasis on objectivity in scientific writing. Concealed sources, such as “It is widely known that…”, appear to be used rather infrequently, though it remains pertinent in knowledge generalization without directly attributing the evaluation to a specific entity. Similarly, evaluation from other individuals, such as “Smith (2020) found that…”, may be limited, as nursing abstracts primarily focus on reporting findings rather than engaging in extensive literature-based argumentation.
Comparing international journal articles (Abbasi Montazeri, Jalilifar, & Hita, 2023) reveals similar trends, with abstract entities emerging as the predominant evaluative source and a limited use of concealed sources, explicit authorial and other-human evaluation. However, studies on scholars in EFL academic settings suggest potential deviations in nursing research abstracts from EFL scholars in Japan. Wang and Chen’s (2012) study of Chinese novice academic writers utilized concealed sources more frequently than established scholars, possibly to reduce overt accountability for their claims.
Building on these insights, we hypothesize that authors publishing in Japanese nursing journals, compared to those submitting to top-tier international journals, may rely on concealed sources to a greater extent. This inclination could stem from challenges specific to EFL academic writing, including limited familiarity with disciplinary conventions and a desire to mitigate the appearance of direct authorial involvement.
In conclusion, although both internationally recognized nursing journal abstracts and Japanese nursing journals make substantial use of abstract entities, the latter appear to employ concealed sources more frequently, reflecting broader tendencies observed in EFL contexts.
2.4. Hypotheses on Evaluative Expression in Nursing Journal Abstracts
Based on findings from medical journals (Kim & Crosthwaite, 2019), we hypothesize that discourse verbs (e.g., “argue,” “state”) will be the most frequently employed the evaluative expressions in nursing journal abstracts, followed by research verbs (e.g., “demonstrate,” “indicate”), nominal predicates (e.g., “the assumption that...”), cognition verbs (e.g., “believe,” “assume”), and adjectival predicates (e.g., “it is important that...”). Experienced authors would presumably make extensive use of discourse acts as a strategic device to assert their positions while mitigating the directness of their judgment (Abbasi Montazeri, Jalilifar, & Hita, 2023). Owing to their familiarity with academic conventions, they could more deftly navigate the delicate balance between conveying evaluative claims with confidence and avoiding potential criticism arising from overt statements of judgment. Furthermore, given that nursing research, like medicine, prioritizes empirical evidence, authors may favor research verbs to present their findings objectively (Shen & Tao, 2021).
In contrast, Hasan and Alsout (2021) reported that authors publishing in EFL contexts tend to employ research verbs more frequently than any other category, while also making notable use of discourse verbs and nominal predicates, yet minimal use of cognition verbs and adjective predicates. A plausible explanation for the predominance of research verbs may be attributed to linguistic and disciplinary conventions that prioritize empirical rigor. Research verbs such as “demonstrate”, “indicate”, and “show” directly link findings to objective methods, bolstering credibility for non-native speakers targeting international academic standards. Moreover, discourse verbs and nouns help maintain a neutral tone and project authority. Notably, nouns often convey a more impersonal and less explicit impression than adjectives, thereby minimizing overt stance-taking and emphasizing factual reporting.
In sum, whereas international high impact factor nursing journal abstracts tend to privilege discourse verbs, with minimal use of cognitive verbs and adjectives, authors contributing to Japanese academic journals may exhibit a stronger reliance on research verbs followed by discourse verbs, with less frequent use of adjectives and cognitive verbs.
3. Methodology
3.1. Data Collection
In accordance with the data collection procedures outlined by Nwogu (1997) and Mwita (2022), journals and abstracts were selected based on the following criteria: 1) international or domestic recognition and high standing in the field of nursing research, 2) publication in English, and 3) inclusion of academic articles with abstracts. This selection process, guided by these criteria, resulted in the collection of a total of 200 nursing academic abstracts, with 100 from each of two journals. The first set was obtained from research articles published in 2024 in international nursing journals with the highest impact factor (7.5) in nursing. The second set comprised 100 abstracts published between 2021 and 2024 in a Japanese nursing journal with the highest impact factor (1.7) among English-language nursing journals in Japan. This selection was because the former is considered the most prestigious international journal in nursing, with articles widely regarded as exemplars of academic writing, while the latter has the highest impact factor (1.7) among English-language nursing journals in Japan as of 2025 and represents the most typical academic writing style found in English-language nursing publications in Japan (see Table 3 for details). Only original research articles were used disregarding reviews, short reports, recommendations, and editorials.
Table 3. Details of the two journal abstracts.
|
International journal |
Japanese journal |
Impact factor |
7.5 |
1.7 |
Abstract |
100 |
100 |
Total number of words |
28,049 |
20,863 |
Average number of words per abstract |
280 |
208 |
Average number of sentences per abstract |
13 |
10 |
Average number of words per sentence |
22 |
22 |
Journal Citation Reports 2023 (Clarivate Analytics, 2023).
3.2. Analysis
This study analyzed evaluative clauses in all abstracts using the framework developed by Hyland and Jiang (2018) and Hyland and Tse (2005), as outlined in the previous section. In order to ensure the reliability of the analysis of evaluative that clauses, the researcher performed a total of three rounds of analysis, each separated by a four-week interval. The first and second analyses differed by approximately 2%. Following a thorough reassessment of the analytical framework and careful scrutiny of ambiguous cases, a third round of analysis was conducted, the results of which aligned completely with those of the second round for all examined items.
3.3. Statistical Procedures
A χ2-square test was employed to examine the statistical significance of the difference in the frequency of evaluative that clauses between abstracts from international and Japanese journals.
4. Results
An analysis of 100 abstracts from the international high impact factor nursing journal identified 49 instances of evaluative that clauses, while 28 instances were found in the Japanese nursing journal (See Table 4). A two-proportion z-test examined whether the difference in occurrence rates was statistically significant. The occurrence rate per 1000 words was 1.7 (49 occurrences in 28,049 words) for the international journal and 1.3 (28 occurrences in 20,863 words) for the Japanese journal. The results indicated no statistically significant difference (z = 1.12, p = 0.26), suggesting that the observed variation may be due to random fluctuations rather than a systematic difference between the two journals.
4.1. Evaluated Entity
The analysis of evaluative that clause entities in the two journals revealed distinct patterns in how evaluations are attributed. The cross-tabulation results are presented in Table 5.
Table 4. Occurrences of evaluative that clause in the two journals.
|
International journal |
Japanese journal |
Number of evaluative that clauses |
49 |
28 |
Number of evaluative that clauses per 1000 words |
1.7 |
1.3 |
Table 5. Evaluated entity distribution of evaluative that clauses in two types of journal abstracts.
|
Author’s findings |
Previous studies |
Methods, models, theories, and hypotheses |
Total |
|
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
International journal |
36 |
73.4 |
7 |
14.2 |
6 |
12.24 |
49 |
100 |
Japanese journal |
28 |
100 |
0 |
0 |
0 |
0 |
28 |
100 |
Total |
64 |
83.12 |
7 |
9.09 |
6 |
7.79 |
77 |
100 |
In both journals, the author’s findings (see Example 1) were the most frequently used entity; however, differences emerged in the distribution of other entity types. In the internationally recognized journal, 73.47% of evaluative statements referenced the author’s findings, while 14.29% attributed evaluations to previous studies (see Example 2) and 12.24% to methods, models, theories, or hypotheses. This suggests that authors in top-tier international academic journals incorporate external sources and theoretical frameworks when making evaluative claims, positioning their research within broader research context. Conversely, the Japanese journal exclusively used the author’s own findings (100%) as the entity of evaluation, with no references to previous studies (0%) or theoretical and methodological frameworks (0%). This indicates a stronger reliance on self-contained evaluations, with limited explicit engagement with previous research or theoretical perspectives.
Example 1. These findings suggest that patients with breast cancer were mainly concerned with… (IJ56Conclusions1)
Example 2. Despite previous studies suggesting that developmental care can provide… (IJ 25Background1)
To determine whether the observed discrepancy in the evaluated entity usage between the two journals was statistically significant, a chi-square test of independence was conducted. The chi-square test revealed a statistically significant association between journal type and entity categorization, χ2 (2, N = 77) = 8.94, p = 0.011. Cramér’s V indicated a moderate effect size (V = 0.34, 95% CI [0.18, 0.57]). Residual analysis indicated that the frequency of the author’s findings was significantly lower than expected in the international high impact factor journal abstracts (z = −2.99, p = 0.003), whereas it was significantly higher in the Japanese journal abstracts (z = 2.99, p = 0.003). Conversely, previous studies appeared significantly more frequently than expected in the international journal abstracts (z = 2.1, p = 0.036) and were completely absent in the Japanese journal abstracts (z = −2.1, p = 0.036). The category of methods, models, theories, and hypotheses did not show statistically significant deviations from the expected frequencies (p > 0.05). These findings suggest that different journal types exhibit distinct preferences in using evaluative that clauses, with Japanese journal abstracts favoring the author’s findings and internationally recognized journal abstracts incorporating a broader range of entities, including previous studies.
4.2. Evaluative Stance
The analysis of evaluative clause stance types in the two journals reveals notable differences in the distribution of affect, obligation, doubt, certainty, and neutral markers. The cross-tabulation results are presented in Table 6.
Table 6. Evaluative stance distribution of evaluative that clauses in two types of journal abstracts.
|
Affect |
Obligation |
Doubt |
Certainty |
Neutral |
Total |
|
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
International journal |
1 |
2.04 |
1 |
2.04 |
35 |
71.43 |
11 |
22.45 |
1 |
2.04 |
49 |
100 |
Japanese journal |
0 |
0 |
0 |
0 |
14 |
50 |
13 |
46.43 |
1 |
3.57 |
28 |
100 |
Total |
1 |
1.3 |
1 |
1.3 |
49 |
63.64 |
24 |
31.17 |
2 |
2.6 |
77 |
100 |
In both journals, doubt (see Example 3) and certainty (see Example 4) were the most frequently used stance types, though their distribution varied. In the international high impact factor journal, doubt markers accounted for 71.43% of cases, while certainty markers comprised 22.45%. Conversely, the Japanese journal showed a lower frequency of doubt markers (50.00%) and a higher occurrence of certainty markers (46.43%). This suggests that the internationally recognized journal favors hedging and cautious claims, whereas the Japanese journal presents more assertive evaluative statements. Affect and obligation markers were minimal in both journals. The top-tier international academic journals contained fewer instances (2.04% for both), while the Japanese journal had none (0%). This suggests that both journals maintain a predominantly objective and impersonal tone, with limited reliance on emotional or obligatory expressions. Finally, neutral stance markers appeared at similarly low frequencies in both journals (international journal: 2.04%; Japanese journal: 3.57%), indicating that purely neutral evaluations are uncommon.
Example 3. A two-way analysis of variance suggested that CM moderated the relationship… (JJ 96Results3)
Example 4. Multiple logistic regression analysis revealed that the independent predictive factors for… (IJ 30Results2)
To determine whether the observed discrepancy in the evaluative stance usage was statistically significant, a chi-square test of independence was conducted. The test did not reveal a statistically significant association between journal type and stance categories, χ2 (4, N = 77) = 5.87, p = 0.21. Cramér’s V suggested a weak to moderate effect size (V = 0.28), though the confidence interval did not converge. Residual analysis indicated that the frequency of certainty markers was significantly lower than expected in the international high impact factor journal abstracts (z = −2.19, p = 0.029), whereas it was significantly higher in the Japanese journal abstracts (z = 2.19, p = 0.029). No other stance category showed statistically significant deviations from expected frequencies (p > 0.05). These findings suggest that while both journals exhibit similar distributions of evaluative that clauses across stance categories, the Japanese journal abstracts tend to favor certainty expressions more than their international counterparts.
4.3. Evaluative Source
The analysis of evaluative that clause sources in the two journals revealed both similarities and differences in their distribution. The cross-tabulation results are presented in Table 7.
Abstract entities (see Example 5) were the most prevalent source type in both journals, accounting for 83.67% of cases in the top-tier international academic journal abstracts and 78.57% in the Japanese journal abstracts. This suggests a shared tendency toward impersonal, concept-driven evaluations. However, differences emerged in the use of concealed sources and author references. The Japanese journal abstracts contained a significantly higher proportion of concealed sources (21.43%, see Example 6) compared to the international journal abstracts (4.08%), implying that evaluations in the Japanese journal abstracts are more frequently attributed to unspecified or implicit sources. Conversely, explicit author references appeared solely in the international journal abstracts (12.24%) and were entirely absent in the Japanese journal abstracts (0%), suggesting that international journal authors are more likely to directly express their evaluative stance, whereas authors who have published in Japanese nursing journals maintain a more detached, impersonal tone. These findings indicate that while both journals predominantly rely on abstract entities for evaluation, the Japanese journal abstracts show a stronger preference for implicit attribution, while international journal abstracts incorporate a more direct authorial presence. The differences in source attribution may reflect distinct rhetorical conventions or disciplinary expectations.
Example 5. The results indicated that ePRO-based symptom management interventions could improve… (IJ91Results1)
Example 6. It was found that the mean central venous pressure value measured… (JJ7Results1)
Table 7. Evaluative source distribution of evaluative that clauses in two types of journal abstracts.
|
Abstract entity |
Concealed Source |
Author |
Total |
|
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
International journal |
41 |
83.6 |
2 |
4.08 |
6 |
12.2 |
49 |
100 |
Japanese journal |
22 |
78.57 |
6 |
21.43 |
0 |
0 |
28 |
100 |
Total |
63 |
81.82 |
8 |
10.39 |
6 |
7.79 |
77 |
100 |
To determine whether the observed discrepancy in the evaluative source usage was statistically significant, a chi-square test of independence was conducted. The test revealed a statistically significant association between journal type and source categorization, χ2 (2, N = 77) = 8.646, p = 0.013. Cramér’s V indicated a moderate effect size (V = 0.335, 95% CI [0.175, 0.567]). Residual analysis showed that the frequency of concealed sources was significantly lower than expected in the internationally recognized journal abstracts (z = −2.400, p = 0.016), whereas it was significantly higher in the Japanese journal abstracts (z = 2.400, p = 0.016). No other source category showed statistically significant deviations from expected frequencies (p > 0.05). These findings suggest that different journal types exhibit distinct preferences in their use of evaluative that clauses, with Japanese journal abstracts favoring concealed sources, while international journal abstracts favor more explicit source attribution.
4.4. Evaluative Expression
The analysis of evaluative that clause expressions in the two journals revealed both shared tendencies and notable differences in distribution patterns. The cross-tabulation results are presented in Table 8.
In both journals, research acts (see Examples 7 and 8) were the most frequently employed category, accounting for 81.63% in the international high impact factor journal and 82.14% in the Japanese journal, suggesting a shared preference for evaluations centered on research-related processes and findings. However, differences emerged in using noun-based expressions, discourse acts, and cognitive acts. Noun-based expressions appeared exclusively in the international journal (6.12%) and were absent in the Japanese journal (0%), indicating that the international journal incorporates more nominalized evaluative structures. Similarly, discourse act expressions were slightly more frequent in the international journal (8.16%) than in the Japanese journal (7.14%), suggesting that explicit evaluations of communicative or rhetorical actions may be more common in internationally recognized journals. Conversely, cognitive act expressions were more frequent in the Japanese journal (10.71%) than in the international journal (4.08%), suggesting that the Japanese journal relies more on evaluation related to mental processes, such as beliefs or interpretations.
Table 8. Evaluative expression distribution of evaluative that clauses in two types of journal abstracts.
|
Noun |
Adjective |
Research act |
Discourse act |
Cognitive act |
Total |
|
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
(n) |
(%) |
International journal |
3 |
6.12 |
0 |
0 |
40 |
81.63 |
4 |
8.16 |
2 |
4.08 |
49 |
100 |
Japanese journal |
0 |
0 |
0 |
0 |
23 |
82.14 |
2 |
7.14 |
3 |
10.71 |
28 |
100 |
Total |
3 |
3.9 |
0 |
0 |
63 |
81.82 |
6 |
7.79 |
5 |
6.49 |
77 |
100 |
Example 7. Our data support the emphasis on the need to retain clinical judgment and suggest that… (IJ 68Conclusion1)
Example 8. Multiple linear regression analysis revealed that illness perception was associated… (JJ 43Results3)
To determine whether the observed discrepancy in the evaluative expressions usage between the two journals was statistically significant, a chi-square test of independence was conducted. The test did not reveal a statistically significant association between journal type and expression categories, χ2 (3, N = 77) = 2.946, p = 0.40. Cramér’s V suggested a weak effect size (V = 0.19), though the confidence interval did not converge. Residual analysis indicated that no expression type showed statistically significant deviations from expected frequencies (p > 0.05), suggesting that both journal types exhibit similar distributions of evaluative that clauses across expression types. These findings indicate no distinct preference for a specific expression type between the two journals.
5. Discussion
This study examined how the hypotheses regarding evaluative that clauses in international and Japanese journals align with empirical findings. The analysis focused on four aspects: the evaluated entity, the evaluative stance, the evaluative source, and the evaluative expressions.
Regarding the evaluated entity, it was hypothesized that international high impact factor journal articles primarily evaluate their own findings with minimal reference to prior research (HI1), whereas Japanese journal articles rely even more heavily on self-evaluation (HJ1). The findings support these hypotheses. In internationally recognized journals, 73.47% of evaluative that clauses refer to the author’s findings, while 14.29% reference previous studies. Conversely, in Japanese journals, 100% of evaluative clauses refer to the author’s findings, with no references to prior research or theoretical frameworks. These results confirm that researchers publishing in Japanese nursing journals engage less with previous studies compared to their international counterparts.
Regarding the evaluative stance, the hypothesis predicted that international high impact factor journal articles predominantly express epistemic doubt, with certainty less pronounced, and attitudinal markers rarely observed. (HI2). Conversely, Japanese journal articles may favor epistemic certainty, place less emphasis on epistemic doubt, and include frequent attitudinal obligation markers (HJ2). The findings partially support these hypotheses. In internationally recognized journals, epistemic doubt is the most frequent stance (71.43%), followed by epistemic certainty (22.45%). In Japanese journals, epistemic doubt occurs less frequently (50%), while epistemic certainty is more prevalent (46.43%), indicating a stronger preference for certainty-based evaluation. However, obligation markers were infrequent in both journal types. These results suggest that researchers publishing in Japanese nursing journals express greater certainty and rely less on hedging than internationally skilled authors. However, the expected prevalence of attitudinal obligation is not evident.
Regarding the evaluative sources, it was hypothesized that international high impact factor journal articles predominantly employ abstract entities as sources, with concealed sources and authorial evaluation playing secondary roles (HI3). In contrast, Japanese journal articles were expected to rely more on abstract sources, make less use of authorial evaluation, and favor concealed sources (HJ3). The findings partially confirm these hypotheses. In top-tier international academic journals, abstract entities constitute 83.67% of sources, while concealed sources account for 4.08% and authorial evaluation for 12.24%. In Japanese journals, abstract entities remain common (78.57%), but concealed sources are more frequent (21.43%), and authorial evaluation is absent (0%). These findings suggest that internationally skilled authors prefer direct attribution, whereas researchers publishing in Japanese nursing journals rely more on implicit attribution.
Finally, the analysis of evaluative expressions considered that internationally recognized journal articles tend to employ discourse verbs and research verbs frequently, while making minimal use of cognitive verbs, adjectives, and nouns. (HI4). Conversely, Japanese journal articles were expected to rely most heavily on research verbs, though discourse verbs and nouns are also employed frequently. However, they seldom make use of cognitive verbs and adjectives. (HJ4). The findings partially support these hypotheses. In internationally recognized journals, research verbs are the most frequent (81.63%), while cognition verbs (4.08%) and noun predicates (6.12%) are comparatively rare. In Japanese journals, cognition verbs are more common (10.71%), but research verbs remain dominant (82.14%). Contrary to expectations, noun predicates are absent (0%). These findings indicate that while researchers publishing in Japanese nursing journals use cognition verbs more frequently than internationally skilled authors, the anticipated reduction in research verbs is not observed, nor is the expected increase in noun predicates.
Overall, the findings largely corroborate the hypothesis that internationally experienced authors and those publishing in Japanese nursing journals exhibit discernible differences in their writing practices, albeit with some deviations. In particular, the strongest evidence indicates that abstracts in Japanese nursing journals rely more heavily on self-evaluation in the use of evaluative that clauses, display a greater degree of epistemic certainty, and favor implicit attribution of sources.
These findings, indicating a distinctive reliance on self-evaluation and concealed sources in Japanese nursing abstracts, accord with prior studies (Lee, 2009; Nozawa, 2015) that underscore the significance of implicit rhetorical structures in Japanese academic discourse. The tendency to highlight one’s own research while minimizing explicit engagement with external sources resonates with cultural norms emphasizing deference to authority and harmonious communication. Such patterns suggest that, even in English-language abstracts, Japanese authors may incorporate culturally embedded writing practices that prioritize courtesy and indirectness over overt critique. Hence, this study confirms and extends existing research on the unique features of Japanese academic writing. Our findings in this nursing-focused study suggest a possible trend, however, whether a similar pattern appears in other academic fields in Japan remains an open question. This underscores the need for further investigation to determine whether our observations reflect a broader phenomenon within Japanese academia.
This study’s findings suggest that the initial hypotheses, particularly those concerning authors writing in EFL settings, were not fully supported. This may be because previous research specifically examining evaluative that clauses in nursing research are lacking. Unlike medicine or applied linguistics, studies on evaluative language in nursing academic writing remain scarce. Previous research, such as Altun (2021) and Abbasi Montazeri, Jalilifar and Hita (2023), focused on applied linguistics rather than nursing. Given the disciplinary variations in academic writing conventions, findings from applied linguistics may not directly apply to nursing research, highlighting the need for further investigation into discipline-specific evaluative practices in nursing discourse.
Recent studies have examined the learning effectiveness of artificial intelligence, particularly tools such as ChatGPT (Li, Lowell, Wang, & Li, 2024). These studies suggest that the researchers publishing in Japanese academic journals analyzed in this study may have benefited from such advancements. This raises the possibility that researchers publishing in EFL contexts have acquired knowledge and skills comparable to those of scholars publishing in internationally recognized journals. Consequently, this may have contributed to the findings, which revealed no significant difference in the frequency of evaluative expressions between authors publishing in Japanese nursing journals and researchers publishing in international high impact factor academic journals.
Notably, the predictions regarding top-tier international academic journals were largely confirmed. This may be from prior research on medical academic writing, such as Kim and Crosthwaite (2019), which informed the hypotheses. Notably, their study focused on full-length research articles rather than abstracts, highlighting that while disciplinary proximity appears to shape evaluative language use, textual genre (i.e., full research articles versus abstracts) may not produce as significant a difference as expected. This suggests that future studies should carefully consider both disciplinary similarities and text-type variations when formulating hypotheses.
6. Conclusion
This study may contribute to the field of academic writing, particularly in nursing research, by examining the use of evaluative that clauses. While prior research on these clauses has been conducted in medicine, their use in nursing has received comparatively limited attention. The findings suggest notable differences from those observed in medical discourse, highlighting the need for further research specifically focused on nursing as a primary subject of inquiry. The results appear to indicate that disciplinary variations in academic writing merit deeper investigation, particularly regarding evaluative language use. This further highlights the need for expanded research on nursing discourse within the broader field of academic writing.
Despite its contributions, this study has limitations. First, the analysis was based on a single international journal and a single Japanese journal, which may not fully represent the diverse writing conventions across different academic publications. Since journals often have distinct stylistic and editorial guidelines, future research should include multiple journals to determine whether the observed patterns hold across a broader range of nursing publications.
Second, the analysis was conducted by a single researcher, which may have introduced subjectivity in categorization and interpretation. Future studies should incorporate multiple analysts to enhance reliability and minimize potential biases.
Additionally, while this study focused on Japanese nursing journals, further research must be conducted in other EFL contexts. The findings indicate that specific trends align with prior research on other academic disciplines, while others diverge, raising concerns regarding whether the observed patterns are unique to Japanese nursing journals or reflect broader tendencies across EFL nursing discourse in other Asian countries. Future studies must consider this to provide a more comprehensive understanding of evaluative language use in nursing research and contribute to the development of academic writing strategies tailored to EFL scholars in the nursing field.
Acknowledgements
This work was supported by JSPS KAKENHI Grant-in-Aid for Scientific Research (C) No. 18K0088. The authors are grateful for their support.