Вы находитесь на странице: 1из 17

Page 1

Report of the Investigation


on
the Chen-Kuan (CK) paper (2017) and the Chang Thesis (2016)

The NATPA Committee on Academic and Research Integrity


March 6, 2018

Data Sources, p. 1
Summary of Findings, p. 2
Introduction: The CK Paper and the Chang Thesis in question, p. 2
Part I Direct Comparison between CK and Chang, p. 3
A. On the Identity between Figures 1, 2 of CK and Figures 2-1, 2-2 of Chang, Respectively, p. 3
B. On the identical and Similar Explanations of the Figures between CK and Chang, p.4
C. On the identity between Table 1 of CK and Tables 4-1, 4-2 of Chang, and between Table 2 of CK and Table 4-3 of
Chang, p. 8
D. On the Identity between Table 3 of CK and Part of Tables 4-8, 4-9, and 4-10 of Chang, p. 9
E. On the References Section, p. 11
F. On Another Kind of Plagiarism, p. 12
G. Some Other Considerations, p. 13
Part II The Questions of Academic and Research Integrity, p. 13
Part III On the Calculation of Growth Rates 成長率 and the Effects of ECFA. P. 15
Acknowledgements, p. 17

Data Sources

One paper and one thesis are compared in this Report.

One is the paper by Jian-Liang Chen and Chung-Ming Kuan (CK) submitted to the Academia Sinica-NTU joint conference
on 2017/04/11, and presented at the conference on 2017/5/6 (see page 2). We will call it CK, or the CK paper in this
Report.

The second paper is MA Thesis by Wei-Ling Chang, 2016/07, which was approved on 2016/07/30 (see page 2). We will
simply call it Chang or the Chang thesis (see page 2) in this Report.

Our investigation was complicated by the existence of another paper by Chen and Kuan cited in the Chang Thesis as 手稿
, but not listed in CK paper. We will call it 手稿 or CK (2016).

Our major task is to determine if violation of research and academic integrity, particularly plagiarism, exits. Plagiarism is
defined as follows:

A. Plagiarism of the first kind: Exact word-by-word copying of another person’s writing.
B. Plagiarism of the second kind: Almost word-by-word copying of the same idea or contents with some change or
arrangement of the wording of other people’s work
C. Plagiarism of the third kind: Taking other people’s idea, design, process, or invention as if they were their own.
Page 2

Summary of Findings:

We show violations of academic ethics and research integrity by the authors of CK paper as evidenced by
plagiarism found in the CK paper and by no citation of the Chang thesis or the 手稿 CK (2016) in the CK paper.

Our findings are based on the universal principles and do not depend on whether the writings are formal or
informal, published or unpublished, term papers or classroom reports.

In addition to the Introduction, this Report is divided into three parts.

Introduction: The CK Paper and the Chang Thesis in question


The relevant part of the title pages of CK paper and Chang thesis are shown below.

The CK paper The Chang Thesis

1. The CK paper acknowledged Chang for providing two data sets as follows

Note that Chang was acknowledged only for providing two publicly available data sources, which by themselves
have nothing to do with the materials in the Chang thesis which CK plagiarized.
Page 3

2. Chang was under the supervision of Chen, and the Chang thesis was approved on July 30, 2016 by Chen and two off-
campus thesis examiners. Chen and his co-author Kuan submitted their paper on April 11, 2017, and presented it on
May 6, 2017, about 10 months after the Chang thesis was deposited.

3. While the titles of CK and Chang appear different, both CK and Chang used the same datasets and identical
“difference-in-differences model” to study the impact of ECFA on the Taiwanese exports to China.

Part I Direct Comparisons between CK and Chang


Identity and similarity between CK paper and Chang thesis were found in at least 16 places. Only some major ones are
shown below.

A. On the Identity between Figures 1, 2 of CK and Figures 2-1, 2-2 of Chang, Respectively

4. Figure 1 of CK is identical to Figure 2-1 of Chang. We reproduce them as follows:

CK, Figure 1, page 36 Chang, Figure 2-1, page 12

5. The "3" of "3A" indicates the 3rd identity/similarity found in the two papers and" A" denotes the type of plagiarism
defined in page 1.

The two figures above share unusual common features. Firstly, neither had a vertical axis title. Secondly, the
positions where the unit of measurement (單位: 百萬元台幣) was placed in both figures are very similar, if not
identical. Most graphic programs require placing such a text box in the figure manually. Third, both choose to
present the same period of trade data, from 1981 to 2014.
In addition, line markers, the legend, the position of the legend, the grid lines, the horizontal and vertical range of
the data, even the wording of the chart title are very similar. The only difference is the font of the horizontal titles of
the CK paper and of the Chang thesis.
Page 4

The similarity observed between these two figures appears to be beyond coincidence. This is consistent with the
identity in description/explanation of these figures which will be shown below.

Although both CK and Chang claimed 作者自行整理, because the Chang thesis was published 9 to 10 months before
the CK paper and because Chang did not attribute Figure 2-1 to CK (2016), Figure 1 of CK should have cited Chang
thesis.

6. Like the above figures, Figure 2 of CK and Figure 2-2 of Chang are also nearly identical, including the position of 單位:
百萬元台幣 and all the graphic elements mentioned in #5. The only exception is that the horizontal title of Figure 2
of CK was truncated by two words. The first line of the CK horizontal title ends at 出超, but Chang’s horizontal title
ends at 出超統計. Both CK and Chang also claimed 作者自行整理.

7. Since Chang completed her thesis which was approved by Chen in July 2016, the CK paper was submitted in April
2017, we conclude that CK paper should have cited Chang thesis. Furthermore, it is highly unlikely that Chang
plagiarized these two figures from the CK paper since Chen would have not approved the thesis. The two-word
difference in horizontal title is likely a copping error in the CK paper.

CK, Figure 2, page 36


Chang, Figure 2-2, page 13

B. On the Identical and Similar Explanations of the Figures between CK and Chang

We next compare how CK and Chang described their figures.

8. In the following boxes, “1A” and “2A” are used to indicate two different parts containing identical sentences. The
underlined words show the identical wording between CK and Chang. They are pervasive and obvious.
Page 5

Note that, the accepted writing practice is that any quotation from other’s work should be short and concise, and should
be quoted with quotation marks, with a complete citation of the source of the quotation. Without these three points,
the identical wording is plagiarism. This basic principle applies universally to all types of writing and does not depend on
whether the reports or papers are formal or informal, published or not published, term papers or not.

Explanation of Figure 1 by CK (2017), page 5, is given as follows:


1
A

Chang explained her Figure 2-1, pages 12, as follows:


Page 6

9. On the other hand, the explanation of Figure 2 by CK and that of Figure 2-2 by Chang are also identical.
The copying is so identical that we can say “not even one character is spared” (一字不漏).

2 Explanation of Figure 2 by CK, page 5.


A

Explanation of Figure 2-2 by Chang, page 13 is as follows:

10. The only difference between the description/explanation of 圖 1 of CK and 圖 2-1 of Chang is that the former starts
with “圖 1 是” while the latter put “參閱圖 2-1” at the end. The corresponding differences are shown by lines with
double headed arrows.

11. Similarly, the only difference between the description/explanation of 圖 2 of CK and 圖 2-2 of Chang is that the
former starts with “圖 2 是” and the latter put the “參閱圖 2-2” at the end. The corresponding differences are
shown by lines with double headed arrows.
Page 7

12. The fact that there is a high degree of similarity between the description of Figure 1 of CK and that of Figure 2-1 of
Chang, and between the description of Figure 2 of CK and that of Figure 2-2 of Chang, as well as the fact that Chang
thesis was approved by Chen and two other thesis examiners about 10 months earlier, indicate that the description
of Figures 1 and 2 were plagiarized from the description of Figures 2-1 and 2-2 of Chang, respectively.

13. Identity and similarity in wording were also found in the conclusions/summary of Figures 1, 2 of CK and Figures 2-1,
2-2 of Chang. To save the space, these are shown side-by-side for comparison below.

CK page 6, first paragraph Chang, page 13.

14. We may point out again that, the identical word-by-word paragraphs between CK and Chang suggest a high
likelihood that CK copied from Chang, since Chang wrote her thesis in 2016 before CK wrote their paper in 2017.

15. Furthermore, since Chang wrote the thesis under the direction of Chen, if Chang had plagiarized the words from 手
稿 CK (2016) and copied the CK ideas in such a direct and extensive way, Chen would have stopped her. More on
this point in #30.
Page 8

C. On the identity between Table 1 of CK and Tables 4-1, 4-2 of Chang, and between Table 2 of CK and Table 4-3 of
Chang

16. #17 to #18 below show that both CK and Chang unambiguously indicated who did the work and the data sources in
the respective footnotes in 資料來源. We list these tables here to show both CK and Chang, especially Chang, was
fully aware of the importance of distinguishing who worked on what and from where the tables or figures were
derived. It should be pointed out that Chen and two other thesis examiners approved Chang's thesis. Chen must
have agreed that the attribution of the sources by Chang for all figures and tables in her thesis was entirely accurate
and appropriate.
CK, Table 1 and 2, page 20 Chang, Tables 4-1, 4-2 and 4-3, pages 24 and 26

17. In Tables 4-1, 4-2, and 4-3, Chang clearly indicated that the tables came from CK (2016), and CK’s footnote indicated
that 作者自行計算.

18. Specifically, in CK Tables 1 and 2, the “資料來源” show


Page 9

And Chang Tables 4-1, 4-2, and 4-3, show:

It is not clear what 原始資料 is, whether it means the data in Table 1 is the original data, or CK aggregated the
government detailed data to compile Table 1 of CK. In any case, the above “引用自陳建良與管中閔(2016)” shows
that Chang was careful in citing her sources.

19. However, we question why CK did not cite his CK (2016) 手稿, like Chang above, in CK Tables 1 and 2 (see Part II
below).

20. What needs to be pointed out though is that, as will be shown in Part III, the calculation by both CK and Chang on
"growth rate" was wrong.

D. On the Identity between Table 3 of CK and Part of Tables 4-8, 4-9, and 4-10 of Chang

21. All the data in Table 3 of CK were copied from the corresponding data in Tables 4-8, 4-9, and 4-10 of Chang. This is
clearly a case of plagiarism. Details are described below:

Table 3 of CK consists of three categories: cumulative total (China and HK) exports, exports to China, and exports to
HK, for the periods 2007 to 2010 and 2011 to 2014, based on the “difference-in-differences (DID) model”
calculation. However, the CK table has no footnote indicating who did the DID calculation.

On the other hand, Chang shows the details of the cumulative values of each category for the first year (2010 and
2011, before and after ECFA, respectively. The same below), second year (2009-2010, 2011-2012), third year (2008-
2010, 2011-2013), and the fourth year (2007-2010, 2011-2014) for three categories, namely, China and HK (Table 4-
8), China alone (Table 4-9), and Hong Kong alone (Table 4-10).

Table 3 of CK and the cumulative values of the fourth year of Chang for Total (Table 4-8), China (Table 4-9), and Hong
Kong (Table 4-10) are identical, as shown by the three arrows below.

22. Note that Table 3 of CK has no sources of the data (資料來源) and no indication of who made the table. By
contrast, all three tables of Chang, Tables 4-8, 4-9, and 4-10, clearly stated the sources of the data and who made
the table:

It is clear that all the data in the Table 3 of CK were derived from Tables 4-8, 4-9 and 4-10 of Chang without
reference. This is a clear case of plagiarism.
Page 10

CK, page 21, Table 3 Chang, pages 34-36, Table 4-8, 4-9.

Chang, pages 36, Table 4-10.


Page 11

E. On the References Section

23. Chang cited eight mass media opinions in Chang Table 2-6 and gave the exact date of publication of the media in her
References section as shown partially below:

Chang, page 18, Table 2-6 (shown only partially)

Chang, page 44, References (shown only partially)

24. In page 3, the third paragraph of CK paper, CK also cited three of Chang’s newspaper reports in the same context
(circled), but gave only the year in parentheses as below.
Page 12

25. While citing the same three newspaper sources might be coincidental, the beginning sentence of the CK above was
almost the same as those from Chang, page 3, first paragraph, as follows:

26. It should be noted that, all the five newspaper citations in CK (2017) were not listed in the CK References section, in
a great contrast with Chang. This also supports the notion that CK copied the newspaper citations from Chang.

F. On Another kind of Plagiarism

27. The example shown below is another kind of plagiarism. The similarity in idea, context, and wording between CK and
Chang, although not word-by-word copying of the whole paragraphs like those mentioned in #8 to #12 above , is
patently clear (This type of plagiarism, unlike the word-by-word copying, which is designated as type "A", is
designated as type "B"). Based on time sequence, the most logical conclusion is that CK (2017) rewrote the
Introduction Section of Chang (2016).

Chang, pages 2 and 3, last and first paragraphs


CK, page 3, 2nd and 3rd paragraphs

28. The above comparison shows that Chang itemized the three problems separately as 1, 2, and 3. But CK combined
the three points without itemization.
Page 13

Note that the last paragraph in CK starts from 關於 ECFA 政策效果, and 三 in Chang starts from 關於 ECFA 執行效果
are the same ones which we already quoted in #24 and #25 (we enclosed them in the corresponding square boxes). We
reproduce them here to show in its entirety, to illustrate the extent of plagiarism and how some type of the plagiarism
was done.

29. Note the slight difference between 政策效果 and 執行效果.

G. Some Other Considerations

30. The possibility that Chang copied the figures and Introduction from CK (2016) 手稿 is highly unlikely for the
following reasons:

(a) As we have seen in #17, #18 and #22 and the entire Chang thesis, Chang did acknowledge her sources in her thesis
very clearly by indicating which part was her own work, which part came from CK (2016) 手稿 and which part
came from sources other than these two sources.

(b) More importantly, Chen was the thesis advisor of Chang. If Chang had plagiarized from CK (2016) in such a
pervasive way, Chen would have asked Chang to cite the sources or to remove the copied materials from her thesis.

(c) In Chang thesis, other than citing CK (2016) at various places when appropriate, there was no trace of thank
you note expressing Chang’s gratitude to Chen. This is very unusual considering the Taiwanese culture on the
relationship between students and teachers; especially, if Chang thesis is largely a copy of CK (2016) 手稿.

(d) It is noted that the CK (2016) 手稿 was cited only by Chang in her thesis, but not by CK in their CK (2017).

(e) Based on the publication records of Chen as of 2013 and Kuan as of 2015 from the Websites, neither Chen nor
Kuan published in the special field of International Economics. It is possible that Chang not only had a significant
role in data collection, explanation, and writing of her own thesis but also part of the CK paper. In fact the
literature review of both the Chang thesis and the CK paper lacked in-depth surveys of the topics on ECFA, which
counted up to several hundred academic papers in Taiwan. This may even suggest that Chang, who was an
academically less trained student, was the author.

Part II The Questions of Academic and Research Integrity


The plagiarism itself is academic dishonest and unethical. In this part, we discuss violations of academic and research
integrity from another angle.

31. CK (2017) wrote in their 摘要,

In light of the earlier thesis work of Chang (2016), it is simply false to claim the originality of being the first to
use the DID model by CK, unless the DID work was described in CK (2016) 手稿 and the CK paper of 2017
Page 14

also cited CK (2016). Thus, knowing the existence of the Chang thesis, the claim is false regardless of CK
(2017) paper is merely an “informal paper” or “pending on revision”

32. The time sequence of the publications involved is important. The CK paper was dated April 11, 2017 and presented
on May 6, 2017, and Chang thesis was dated July 2016. When CK submitted their paper, Chang had already finished
her thesis with Chen as her adviser. Knowing the existence of the Chang thesis that deals with the same topic, using
the same data set, applying the same method of investigation, and had similar results, the CK had the obligation to
discuss the strengths and weaknesses of Chang paper, how the CK paper differs from the Chang thesis and any novel
contributions of CK paper to the literature. In this regard, not citing Chang thesis in the CK paper by itself is
intellectually dishonest and academically unethical, even without considering the word-by-word plagiarism listed in
Part II.

33. Kuan’s coauthor Chen claims “研究生引用他的部分研究手稿,爆料者 ‘倒因為果.’” (see 管中閔 维基百科,自由
的百科全书. Downloaded 2018/02/26), implying that it was Chang who plagiarized from the CK paper (2016).

This claim was not supported by our investigation. First, “研究生引用他的部分研究手稿” was duly acknowledged
in Chang thesis. Furthermore, Chang thesis was approved by Chen and two committee members and published in
2016. Second, legally, the copyright belongs to Chang. Therefore, CK in 2017 had the obligation to quote Chang
thesis and discuss the pros and cons of Chang’s research results. This does not depend on whether CK 2017 paper is
formal or informal, published or unpublished. Not doing so and claiming Chang’s research results as their own is a
copyright infringement. Both Chen and Kuan are thus judged intellectually dishonest and academically unethical.

34. Chen’s claim of “研究生引用他的部分研究手稿” raises another very important question. CK’s Table 1 and 2, in CK
(2017), page 20, CK wrote (we reproduce #18 for clarity here):

But, Chang attributed Tables 4-1, 4-2, and 4-3, corresponding to Tables 1 and 2, to:

These clearly indicate that the source of CK Tables 1 and 2 was in 手稿 CK (2016). Therefore, the footnote of
their Tables 1 and 2 should have quoted themselves, exactly like Chang as follows:

Or, CK (2017) should have quoted Chang thesis, because the Chang thesis was completed about 9 to 10 months
earlier.

When CK attributed Tables 1 and 2 of CK paper to 作者自行計算 without citing CK (2016) 手稿, CK’s claim that they
were the first one to use DID to study the impact of ECFA on Taiwan exports (see #31) should be judged as a
violation of academic and research integrity because Chang thesis containing the same topic, data and model was
published 9 to 10 months earlier.

35. Lastly, the “growth rate” presented in Table 2 of CK was not the growth rate defined in the field of economics (see
Part III). It is surprising that such a glaring error was not picked up by Kuan, unless Kuan’s claim that he did not know
the existence of CK paper was true. In this regard, it is noted that Kuan was listed as a co-author of Chen in at least
13 previously published papers (See Chen’s publications in National Chi Nan University website. Downloaded on
2018/02/25). Thus, the CK paper is not the only paper in which Kuan was listed as a co-author. In fact, most of
Page 15

Kuan’s other papers were also co-authored. In the realm of academic and research integrity, being a co-author of a
paper means that one has made a real contribution to the findings of the given paper. Further investigations will be
required to know the exact role of Kuan in the CK paper.

Part III On the Calculation of Growth Rates 成長率 and the Effects of ECFA
On the calculation of the growth rates in Table 2 in CK, or Table 4-3 in Chang, the two tables are identical, and they
were calculated from Table 1 of CK or Table 4-1 of Chang.

36. The titles of these two tables were identical,

成長率 here is wrong. In fact, each entry of CK’s Table 2 is merely the export value, n, of Taiwan’s exports to China
and/or HK, normalized by mid-period average value, a, of export values. They are normalized export values, n/a,
NOT growth rates. This is very elementary. Since Chang attributed her Table 4-3 to CK (see #18), whether the CK
paper is formal or informal, considering the status and background of Chen and Kuan, this error is indeed amazing
and incredible.

37. CK (and Chang) also tried to derive their “growth rates” by drawing the figures. Their figures also did not tell us
growth rates. Those figures only showed the trend of the normalized export values.

Examining Table 2 in CK showed that after setting up the normalized table, CK (and Chang) then tried to present
their “growth rates” of exports before and after ECFA using figures. CK (and Chang, who seemed to be confused) did
not give the formula they used. However, we can easily detect the formulas they used from CK’s description in the
second paragraph of page 8. Their “growth rate” calculation is simply

1-(n/a) = (a-n)/a before ECFA and

n/a – 1 = (n-a)/a after ECFA.

These formulas have nothing to do with growth rates.

38. The correct formula for the (discrete) growth rate should be

(nt – nt-1)*100/nt-1, where nt is the value at time t and nt-1 is the value at time t-1.

Table X below shows the correct growth rates of export values based on CK Table 1 in #16, which is reproduced and
placed above Table X for reference
Page 16

39. The exports growth rates of Total Exports, non-ECFA Goods, and ECFA Goods to China and Hong Kong (C+H), China
alone (C), and Hong Kong alone (H), respectively, are illustrated in the following three diagrams.

Export Growth to China and HK, Total


50 Gorwth Rates, 2007-2014
40 Before ECFA After ECFA
30 H
Percent (%)

20 C+H
10

0
C

-10

-20
2007 2008 2009 2010 2011 2012 2013 2014
Year
C+H C H
Page 17

40. We can tell many stories about the above three diagrams. In any case, so far as the aggregate data of Table 1
indicates, it is clear that, in all cases, the export growth rates for China and Hong Kong decreased drastically after
ECFA was implemented in 2011, contrary to assertion by CK and Chang. But this is not the purpose of investigation
here.

-------------------------------
Acknowledgement:

In writing this Report the NATPA Committee of Academic and Research Integrity received supports from the NATPA
members, some data are provided by people in Taiwan. Their comments and suggestions for the betterment of
academic and research environments of motherland are greatly appreciated.

Вам также может понравиться