Introduction

Iranian research and academic centers have provided information databases since the 1990s, and now, by providing electronic databases, Iranian academic libraries try to provide faculty members and students with the latest scientific information (Jowkar and adaehghani, 2006, p. 83). The importance of this issue is double in the field of medicine and related matters, which deals with the health of humans, in addition to educational, research and creating content issues, the correct and principled treatment of patients is also possible using and studing updated information. The databases that universities share in and provide to their researchers are one of information systems types that can be studied the use rate of them. Nowadays, the collection creation program of many libraries and research centers involves purchasing electronic resources and sharing relevant resources. Meanwhile, the studies of users of these electronic resources are considered as necessary for libraries and research centers in terms of the frequency of use, the reasons for using these resources, the types of use of these resources, the views and thoughts of the user on the sources, types of the use of these resources, the opinions of users about these resources, the rate and reasons of user satisfaction, unmet needs and demographic data. Voorbij and Ongerin (2006) believe that only user studies can reveal the motivations, beliefs, aspirations and problems of users in the use of information resources. This study is necessary to help the officials of Shahid Beheshti University's information centers to adopt policies for choosing and providing electronic resources, including databases based on users' demand.

This study aimed to determine the used rate the shared databases by faculty members and graduate students of Shahid Beheshti University. The sub-objectives of this study are:

1. Identifying the use rate the shared databases by faculty members and graduate students of Shahid Beheshti University and determining the reasons for using these databases by these individuals.

2. Determining the rate of the user's success in interacting with the Shahid Beheshti University's shared databases and identifying the strengths and weaknesses of access to these databases in response to user's needs.

3. Determining the estimated information needs of faculty members and graduate students of Shahid Beheshti University by Shahid Beheshti University's shared databases.

4. Identifying the main motive to search information in Shahid Beheshti University's shared databases by faculty members and graduate students of this university.

5. Identifying the restrictive factors in the optimal use of Shahid Beheshti University's shared databases by faculty members and graduate students of this university.

6. Identifying the attitudes of faculty members and graduate students of Shahid Beheshti University about the need to train the use of Shared Beheshti University's shared databases.

7. Identify the rate of faculty members and graduate students' satisfaction or dissatisfaction of Shahid Beheshti University from shared databases of this university

8. Determining the most used and least used shared database of Shahid Beheshti University.

Several studies have been conducted on utilizing the shared electronic resources. In a study titled "investigating the use rate of Ankara University's faculty members from the digital library, Atilgan, D., & D. Bayram (2006) concluded that most faculty members are aware of the shared information sources. Their research showed that the most used databases of Ankara University's faculty members were ISI, Ebsco, ScienceDirect; also the research was most important reason for using these databases. The main reasons of respondents in the lack of use of databases were unawareness how to use the databases and meet the information needs through other sources. Angel Borrego et al. (2007) reviewed the faculty members in the Consortium of Catalonia University Library for the use of electronic journals. Aboue 95% of studied individuals stated that they are very familiar with electronic journals, and 52% of them mainly use electronic journals. The main reasons of those who did not use electronic journals were the lack of familiarity with these journals and the lack of electronic journals with association with their disciplines. In response to the preference of electronic journals on printed journals, 76% of respondents expressed that they preferable the electronic version if there are both print version and electronic version of a journal. Kayaoglu (2008) reviwed the use of electronic journals by the faculty members of Istanbul University, Turkey. Most of the participants in his study expressed that they use e-journals a lot; he also found that there is a significant relationship between the use of e-journals and the field of study of individuals. The users of these resources, who were the student of basic science and medical, used them for training, flipping, research and updating themselves, and the students of human sciences did not use the resources. From the faculty members' view, the most important advantages of e-journals in comparison with printed journals included the twenty-four-hour access to these journals, no need to attend in library, accessing to past numbers, being meta-text of journals, and accessing the more up-to-date versions. Trivedi, M., & A. Joshi (2009) found in a study titled "the use of e-journals versus printed journals by medical researchers at the HM Patel of Medical Sciences, that most researchers use e-journals as much as print journals. This study showed major problems in the lack of optimal use of e-journals, the lack of computer skills, the inability to retrieve accurate and appropriate medical information through various databases such as ScienceDirect, Ebsco, ProQuest, etc.; also, the technical defects posed by this study are reasons for non-use of e-journals such as low download speed, slowness of connection and physical pressure on the eyes. Ahmed (2013) has studied the rate of students' use and satisfaction with shared e-resources in two specialized universities in Bangladesh. This study showed that the main problems for accessing the students to e-resources are limited access to computers and low download speed, which is related to the weak information and communication infrastructure, as well as the unwillingness of students to regularly use these resources. MaxData (2007) used U.S.I. researchers' log file analysis method to determine the rate of use of Ohiolink database by academic users. The findings of this study showed that there are differences between searching information and its usage in universities and research institutes, which, to a large extent, are subject to research activities and the number of users.

Latifi and Osareh (2010) concluded in a study titled 'Identifying the information needs of graduate students at Bu-Ali Sina University, Hamadan, by emphasis on the available databases in this university includes Rose-net Digital Library and Daneshyar Bases and using the questionnaire, that the main purposes of the students are to search information, find out the background for doing research and study projects. Among the existing e-databases, the importance of Elsevier/ScienceDirect is higher (89.2%), afterward Springer, ISI, Scopus, ProQuest, ACS, Oxford Journals, Emerald, ASME, Ebsco, ASCE, IOP and AMS databases are placed in second to thirteenth positions.

Khalili and Matlabi (2010) examined the faculty members' the use and satisfaction with the subscribed databases of Urmia University of Medical Sciences. Using a questionarre, they concluded that the use rate of these bases varies among faculty members, and IranMedex databases has the most usage (13.6%) among the university's subscribed databases, and the Timeh database has the lowest usage (1.4%). The satisfaction with databases is in average level based on charactristics such as being up-to-date databases, the number of databases, the adaption of databases with specialized needs, and the quality of bases' information. The main purpose of using databases is to conduct research activities (21.4%). The communication lines' low speed with (23.6%) is the most serious problem and the lack of familiarity with foreign languages (2.2%) is the least importance Therefore, according to the role of databases or e-journals in promoting the updated information of academic users and publishing scientific research at the university and at the national and international levels, this thesis investigates the use rate of shared databases by the faculty members and graduate students of Shahid Beheshti University to provide solutions to complete existing studies.

Research method

The current study was applied in terms of purpose and was survey-webometrics in terms of method. The population consisted of the 780 people of faculty members and 2,300 people of graduate students of Shahid Beheshti University. Sampling method was a stratified random method and the sample size was determined using Morgan sampling table. Shahed Beheshti University's studied shared databases in this study included:

Farsi databases like Civilica and Magiran and Latin databases like ACM, ACS, APS, ASCE, ASME, Cambridge jornals, EBSCOHOST, Emerald, IEEE, SAGE, ScienceDirect, Scopus, Springer, web of science, WILEY online library, IOP, JSTOR, MathSciNet, OXFORD, RSC.

The sample size consisted of 330 people of graduate students and 266 people of faculty members. In this study, to cellect data, we used two methods of distributing the questionnaire and analyzing the web report. We used a researcher-made questionnaire to collect data. After assessing this questionnaire by some experts of knowledge and information science confirming its validity, we calculated the reliability of the questionnaire by Cronbach's alpha coefficient, which was 0.83. To analyze the data, we used descriptive and inferential statistics such as frequency and percentage, Chi-square and Friedman tests. The SPSS software was used in this study. In order to investigate the use rate of the Latin and Farsi shared databases in Shahid Beheshti University, we used the log files of these databases were obtained from the central library server.

Findings

Table 1 illustrates the frequency and percentage of graduate students. There are 87.5% of students with master's degree.

Answer to the first question of research

1. To what extent do the faculty members and graduate students of Shahid Beheshti University use shared databases?

As shown in Table 4, the first option (very low) has been selected by 65% of the faculty members and 13.3% of the students selected.

According to the results of Table 5, it can be concluded that there is a significant difference between the expected frequency and frequency of observing the variable of the use rate of shared databases in the sample (both faculty members and graduate students).

As seen in Table 6, the access to up-to-date information is the most important reason to use shared databases by faculty members and graduate students.

Answer to second question of research

How much is the rate of successful of user interact with shared databases?

Table 7 presents the relavant frequency and percentage of each selected option by faculty members and graduate students. The first option (very low) has been selected by 3.10% of the faculty members and 6.7% of the students.

Answer to the third question of research

To what extent are the needs of faculty members and graduate students of Shahid Beheshti University met by shared databases?

Table 8 illustrates the relevant frequency and percentage of each option selected by faculty members and graduate students. The option of (very low) has been selected by 33.6% of the faculty members and 33.3% of the students.

According to the results of Table 9, it can be concluded that there is a significant difference between the expected frequency and frequency of observing the variable of the use rate of shared databases in the sample. The option (higher t5han moderate) has been selected by faculty members and graduate students.

Answer to the fourth question of research

What is the main motive to search information in shared databases by faculty members and graduate students of Shahid Beheshti University?

As shown in Table 10, conducting research projects and patent and exploration are the most important motivation to search information in shared databases faculty members. Helding and attending workshops are the most important motivation of graduate students to search information in shared databases.

Answer to the fifth question of research

What are the restrictive factors in the optimal use of shared databases by faculty members and graduate students of Shahid Beheshti University?

Table 11 shows that unfamiliarity with databases and how to search are the most important restrictive factor of faculty members for optimal use of shared databases, and for graduate students, unawareness of the existence of databases is the most important restrictive factor.

According to the results of Table 12, it can be stated that there is a significant difference between the ranks average of the mentioned reasons in the sample (faculty members and graduate students).

Answer to the sixth question of the research

What is the attitude of faculty members and graduate students of Shahid Beheshti University towards the need for training to use the shared databases?

As shown in Table 13, from the faculty point of view, to get help from colleagues is the most important method to train the optimal use of database, and from the graduate student's view is the training method by software packages or reading guides and brochures.

Answer to the seventh question of research

What is the degree of satisfaction from the Shahid Beheshti University's shared databases by faculty members and graduate students of this university?

Table 14 shows the relevant frequency and percentage of each option selected by faculty members and graduate student. The fourth option (high) has been selected by 28% of the faculty members and 15.8% of the students.

Table 15 shows that there is a significant difference between the expected frequency and frequency of observing this variable in the sample (faculty members and graduate students).

Answer to the eighth question of research

How are the use rate of Farsi and Latin shared databases in Shahid Beheshti University?

To answer this question, we used log files from the central library of Shahid Beheshti University. The studies showed that the rate of loaded articles from the Civilica database during the first semester of 2014 to the first semester of 2015, were 1394 and 25672 articles, respectively, and was 32094 articles for the Magiran base.

Increasingly, it determines that, among the Latin databases, ScienceDirect has most used with 13,362 downloads, and among the Farsi databases, Magiran has most used with 32094 downloads; also among all of the Latin and Farsi databases, Magiran has most used. Among the Latin databases, MathSciNet has least use with 163 downloads. Among the Farsi databases, Ciavilla database with 25,672 downloads has the lowest use compared to Magiran database.

Discussion and conclusion

About the satisfaction rate of the respondents from Shahid Beheshti University's shared databases, the results showed that the satisfaction rate was high and at the desired level. In a tudy on the faculty members of Iran's medical sciences universities from specialized databases, Salajegeh (2010) expressed there is a high satisfaction rate from databases. The users' satisfaction from databases or e-journals may be related to various reasons, including the ease of use, the quality of information, the database information related to the requested subject, the accuracy of search, up-to-date information, or perhaps the suitability of their number. The log files of Farsi and Latin databases (as the number of articles downloaded from the databases as PDFs) showed that ScienceDirect has been most used among the Latin databases, Magiran has been most used among the Farsi databases; also, Magiran has been most used among all the Latin and Farsi databases. MathSciNet has been less used among the Latin bases and Ciolica has been less used compared to Magiran among the Farsi databases. Several databases are provided by universities and research centers (depending on their activities) for the user community. The huge volume of provided databases in the world, these centers are forced to select by financial constraints, and the facilities of universities and research centers. In fact, the base selection is based on the scientific needs of the user community, the comprehensiveness, quality and reliability of the information, as well as to be update the information in them (Khalili, 2013, p. 27). Khaseh and Hatami (2005) showed the positive attitude of the faculty members of the faculty of human sciences of Shiraz University towards the databases such as ScienceDirect, Elsevier, and ISI, which are respectively the most used databases by faculty members of human sciences of Shiraz University. According to Atakan, C, et al. (2008), the most used databases by faculty members at Ankara University were: WEB OF SCIENCE, ScienceDirect and Ebsco. In a study on faculty members at Ankara University, Atilgan and Bayram (2006) concluded that the more popular databases of these individuals were ISI, ABC, ScienceDirect. Davar Panah (2010) also expressed, among its six studied databases, Elsevier has the highest use rate, and Oxford has the lowest use rate. In Rajabi's research (2009), Jstor database placed in highest position in terms of the number of downloaded articles. In a research on graduate students at Bu-Ali Sina University in Hamedan, Latifi and Osareh (2010) concluded that the significance of the ScienceDirect and Elsevier database is higher among the existing electronic databases. The result of this section of the study that the ScienceDirect has the highest use rate in comparison with other bases is consistent with the results of Khaseh and Hatami (2005), Atilegan and Bayram (2006), Latifi and Osareh (2010) and, in part, with the results of Hayati and Hassan Shahi (2008), Atakan et al. (2008). It should be noted that Slmani Nadooshan et al. (2008) and Latifi and Osareh (2010) used the questionnaire to determine the use rate of databases. We used the log files in this section of the study; also, the results of this section of the study are not consistent with the results of Davar Panah (2010) and Rajabi (2009). This may be due to the lack of uniformity or diversity of shared databases in each university; also, it may not be reasonable to compare in this case, since selecting e-journals in each university depends on the size of each university's budget and its policy, but ScienceDirectories database has the high use rate among all the databases and in most of the previous studies. It can be guessed accurately using the log files to determine the bases with more and less usage rather the questionnaire; because it is clear that users cannot accurately determine in the questionnaire which bases are used less or more, but can not be stated that the log files are perfect because the log files show the number of download articles as PDF and it is not possible to determin the use rate of these databases just that the articles have been downloaded from the databases.

The current study showed that the most users of the students are the students of master's degree with 87.5%. The results of Tenner, E. and Z.Y. (Lan) Yang (1999) showed that faculty members with associate professors used e-journals more than faculty members and professors. The results of Salajegeh (2010) are based on the use of e-journals and specialized databases, and the existence of a relationship between scientific rank using information centers. The results indicate that there is no difference between professors, faculty members and professors about the use of ejournals, and all three groups are the users of these resources. These results were consistent with the results of Tenner and Yang (1999), but did not consistent with the results of Salajegeh (2010). In this study, most users (10.3% of faculty members) are computer science specialist. Kayaoglu (2008) reviewed the use of e-journals on faculty members at Istanbul University in Turkey. The users of these resources were more the student of basic science and medical, used them for training, flipping, research and updating themselves, and the students of human sciences did not use the resources. In the age of technology and information, computer science, like medical disciplines, is encountered fundamental change; so, researchers in this field need new information in their area of expertise. Hence, databases can meet this that allow the use of new information in e-journals for individuals. In contrast, the law faculty has least used databases with a score of 5.0%. One of the possible reasons for this may be the same reason about students of human sciences, which may be due to the excessive attachment of expertise or discipline to relevant previous sources. Thus, this study aimed to investigate the use rate of Shahid Beheshti University's shared databases by faculty members and graduate students of this university that partially satisfies the need of this research community, i.e., in this study, the important indicators of measuring the usefulness of databases, including the use rate and satisfaction of databases, the degree of satisfying the users' needs by databases, and the rate of users' success for interacting with these bases, are in moderate level in term of users (not at a low level or at a high level). Except ScienceDirect and Magiran and Civilica databases, most databases are used slightly, one of the possible reasons for this is the comprehensiveness of these databases or becoming more known by the educational system. These databases (less-used bases) primarily require informing by librarians and the knowledge and information science's experts and training by academic educational centers through experienced professors; otherwise, it is necessary to disconnect these databases and replace the more appropriate databases. Such an action will increase the number of readers and enhance their use rate; also, it will greatly decreased the university costs. It should also be noted that quantity often is more reffered quality, for this purpose it possible to make an alternative policy to cost-effective databases; i.e., it can be replaced the costs of subscribing in a number of least-used databases with training the use of useful databases, that is, the quality is replaced the quantity.

References

Ahmed, S.M.Z. 2013. A survey of students' use of and satisfaction with university subscribed online resources in two specialized universities in a developing country. Library Hi Tech News. 3(30), 6-8.

Atakan, Cemal. 2008. An evaluation of the second survey on electronic databases audience digital libraries. Usage at Ankara University Digital Library. The Electronic Library 26(2): 249-259.

Atilgan, Dogan & Bayram, Ozlem (2006). An Evaluation of faculty use of the Digital library at Ankara University, Turkey. The journal of academic librarian ship, 32(1).

Covey, D. T. 2005. Using data to persuade: State your case and prove it. Library Administration & Management, 19(2):82-89.

Dilek-Kayaoglu, Hulya. 2008. Use of Electronic Journals by Faculty at Istanbul University, Turkey: The Results of a Survey. Journal of Academic Librarianship. 3(34). 239-247.

Jowkar,AR and Dehghani, L .2005. The study of the use of postgraduate students in educational and psychology faculties of Ferdowsi University of Mashhad, Shahid Chamran of Ahvaz, Tehran, Isfahan and Shiraz from electronic journals compared to printed journals. Studies in Education and Psychology. 7(1): 83-103.

Kaur, B., & R. Verma. 2009. Use of electronic information resources: A case study of Thapar university. DESIDOC Journal Of Library & Information Technology. 29 (2). 67-73.

Khalili, M & Matlabi, D .2014. Online Services at the Urumia University of Medical Sciences: a use and user study. Journal of National Studies on Librarianship and Information Organization. 24 (4): 24-40.

Latifi, F and Osareh, F .2009. Identifying the Information Needs of Postgraduate Students at Hamedan University, with emphasis on available databases in this university. Epistomology, 3(9): 75-63.

Rajabi, A .2008. A survey on using e-journals database in Imam Sadegh University, Research on Information Science and Public Libraries, 15 (2): 193-216.

Salajegeh, Mozhdeh. 2010. Assessment of satisfaction from information centers, electronic journals and specialized databases, and their relationship with age, gender, and academic rank of faculty members. Science and Information Technology, 25(3):431-447.

Tenner, E. and Z.Y. (Lan) Yang. 1999. End-User Acceptance of Electronic Journals: A Case Study from a Major Academic Research Library. Technical Services Quarterly. 17(2): 1-11.

Trivedi, M., & A. Joshi. 2009. Usage of electronic journals (e_journals) versus print journals by health care professionals in HM patel center for medical care and education (HMPCME). [electronic version]. Journal of Health Information in Developing Countries.

Neda Dehghani

MA of Information Science and Knowledge

Nedadehghani18@gmail.com

Amir Reza Asnafi

Faculty member of Information Science and Knowledge Department, Shahid Beheshti

University, Velenjak, Tehran, I.R. Iran, aasnafi@gmail.com

Mohsen Hajizeinolabedini

Faculty member of Information Science and Knowledge Department Shahid Beheshti

University, Velenjak, Tehran, I.R. Iran, zabedini@gmail.com

Neda Dehghani

Shahid Beheshi University, Nedadehghani18@gmail.com

Amir Reza Asnafi

Shahid Beheshti University, aasnafi@gmail.com

Mohsen Hajizeinolabedini

Shahid Beheshti University, zabedini@gmail.com

Table 1. Frequency distribution of students based on the variables of
the grade

Academic Rank        Levels        Frequency      Percentage

                 Master of Art        289           87.5%
                     Ph.D.             41           12.5%

Table 3--Frequency distribution of samples based on the academic rank
of faculty members

Academic           Levels            Frequency    Percentage
Rank

                 Instructor              22            8.4%
             Associate Professor         63           24.3%
             Assistant Professor        170           65.4%
                  Professor               5            1.9%

Table 4: Stratified frequency distribution of the use rate of shared
databases by faculty members and graduate students

             Frequency                 Percentage

Options      Professor    Student      Professor    Student

Very low     40           44           65%          13.3%
Low          56           50           21.5%        15%
Moderate     100          124          38.3%        37.5%
Much         70           88           27.1%        26.7%
Too much     17           25           6.5%         7.5%
Total        260          330          100%         100%

Table 5--Inferential statistics indicators used for analyzing the Chi-
square test

population           Number    [chi square]

Faculty Members        260         80.3
Graduate students      330         103.8

population            Degrees of    Significance
                       freedom         level

Faculty Members           4            P<0.01
Graduate students         4            P<0.01

Table 6. Stratified frequency distribution of reasons for using shared
databases by faculty members and graduate students.

                                                    Ranks Average

Reasons                                        Professor     Student

Availability of e-resources                       3.79         3.95
Quickly searching and retrieving information      4.07         4.04
Access to up-to-date information                  4.61         4.47
No search restriction if others use the base      3.84         3.63
Feasibility of saving required information        4.12         4.32
  on the computer
High quality electronic resources                 3.72         3.81
Not having time and space restrictions on         3.85         3.79
  receiving electronic resources

Table 7. Stratified frequency Distribution of studied variable in
faculty members and students

                    Frequency                 Percentage

Options      Professor     Student     Professor     Student

Very low         27           22         10.3%         6.7%
Low              56           58         21.5%        17.5%
Moderate         90          171         34.6%        51.7%
Much             70           72         27.1%        21.7%
Too much         17           8           6.5%         2.5%

Total           260          330          100%         100%

Table 8. Stratified frequency Distribution of studied variable in
faculty members and students

                    Frequency                 Percentage

Options      Professor     Student     Professor     Student

Very low         24           14          9.3%         4.2%
Low              34           25         13.1%         7.5%
Moderate         95          162         36.4%        49.2%
Much             87          110         33.6%        33.3%
Too much         19           19          7.5%         5.8%
Total           260          330          100%         100%

Table 9. Inferential statistics indicators used to analyze the Chi-
square test

Example             Number    [chi square]   Degrees of   Significance
                                              freedom        level

Faculty Members       260         82.9           4           P<0.01
Graduate students     330        294.5           4           P<0.01

Table 10--The ranks average of studied variable in faculty members and
graduate students

Ranks Average

The motivation to search information in        Professor     Student
shared databases

Obtaining specialized information                 3.99         4.40
Writing and translating books and articles        4.66         4.46
Enhanceing personal and professional              4.48         4.57
  information
Preparation for teaching and improving the        5.06         4.69
  teaching quality
Participating in domestic and foreign             4.75         4.94
  conferences
Maintaining professional and specialized          5.31         5.19
  competence
Holding and attending in workshops                5.67         5.74
Executive tasks                                   5.35         5.42
Conducting research projects and patent and       5.73         5.60
  exploration

Table 11. The ranks average of the studied variables in faculty
members and graduate students

Ranks Average

The restrictive factor in the optimal use of   Professor     Student
shared databases

Unawareness of the existence of bases             4.87         5.05
Non-relevance the existing information in         4.82         4.77
  the database to the information needs
Unfamiliar with the bases and how to search       5.23         5.03
time limitation                                   4.43         4.32
The lack of access to past ages                   4.70         4.61
The lack of feeling need to use bases             3.04         3.29
The lack of easy access to the internet           5.19         4.39
Unfamiliar with foreign languages                 3.75         4.55

Table 12--Inferential statistics indicators used for Friedman test
analysis

Variable            Number    [chi square]   Degrees of   Significance
                                              freedom        level

Faculty Members       230        169.6           7           P<0.01
Graduate students     360        162.6           7           P<0.01

Table 13--The ranks average of the studied variables in faculty
members and graduate students

                                                    Ranks Average

Methods of training the best use of            Professor     Student
databases

Attending in classrooms and training              2.52         2.88
  workshops
To get help from librarians                       2.81         2.97
Training by software packages                     2.66         3.08
Reading guides and brochures                      3.00         3.06
To get help from colleagues or classmates         4.01         3.02

Table 14. Stratified frequency distribution of the use rate of shared
databases by faculty members and graduate students

                     Frequency                Percentage

Options      Professor     Student     Professor     Student

Very low         34           36         13.1%        10.8%
Low              29           33         11.2%         10%
Moderate        122          198         46.7%         60%
Much             73           52          28%         15.8%
Too much         2            11          0.9%         3.3%
Total           260          330          100%         100%

Table 15. Inferential statistics indicators used for analyzing the
Chi-square test

Sample              Number   [chi square]   Degrees of    Significance
                                              freedom        level

Faculty Members      260        135.6            4          P <0.01
Graduate students    330        374.3            4          P <0.01

Table 15. The use rate of Latin shared Databases at Shahid Beheshti
University

Database name           The rate of Loaded articles

MathSciNet                          163
ScienceDirect                      13362
Springer                            8062
WILEY online library                5052
Scopus                              5378
JSTOR                               4863
EBSCOHOST                           4241
Emerald                             3394
web of science                      2285
IEEE                                2191
OXFORD                              1653
Cambridge                           1501
SAGE                                1248
ACS                                 934
IOP                                 595
ASCE                                569
ACM                                 424
APS                                 402
RSC                                 381
AIP                                 357
ASME                                188