Developing and validating translators’ book selection criteria scale: The case of translators in Iran

: The present study aimed to develop and validate an instrument to measure Iranian translators’ book selection criteria (TBSC). A mixed-methods approach was adopted to conduct the research. During the qualitative phase, the researchers developed the first draft of the questionnaire by generating items through focus groups and literature review. This led to constructing a 40-item closed-ended questionnaire ranging on a 6-point Likert scale from 1 ( strongly disagree ) to 6 ( strongly agree ). Participants of both phases were selected by snowball sampling and participated voluntarily in the study. The quantitative approach was used to pilot the web-based questionnaire with 50 participants. To assess the psychometric properties of the final instrument, a sample of 251 translators completed the questionnaire. The revised and modified version of the TBSC scale included 34 items loaded on six subscales, namely general considerations, external motivating factors, success guarantee factors, translator’s authority/agency, internal motivating factors, and finally, publisher’s authority/agency. The study’s findings indicate that the TBSC is a valid and reliable (α = .748) instrument to measure the criteria affecting the translator’s book selection in the context of Iran.


Introduction
Humans have always encountered situations where they have had to make choices in their personal, social, and professional lives. Ostensibly, each selection will entail rejecting other available options based on a rationale or some type of reasoning. The translator as an individual and a social agent is not an exception. The moment a translation is commissioned, the translator will be faced with a set of choices. The first major decision to be made is whether to accept a translation project. In the next stage, the translator must decide on and apply the appropriate strategies for translation. Literature abounds with discussions on translation strategies for different text types (e.g., Chesterman, 1997Chesterman, /2005Hurtado Albir, 2002;Koster, 2002;Lorscher, 1991Lorscher, /2005.
However, there is a shortage of literature addressing translators' criteria for text selection. The lack of studies on the first step of any translation project, i.e., material selection, encouraged the authors of this paper to conduct a study to identify influential factors that hinder or promote the selection or non-selection of a book for translation in the context of Iran.
To find the Iranian translators' book selection criteria, the researchers decided to develop a questionnaire. This decision was taken on two grounds. First, at the time of this study, no pre-designed questionnaire was available to investigate this issue. Second, it is reasonable to assume that, at least, some of the reasons for choosing a book for translation may vary from one society to another. In other words, the conditions of publication, interest, taste, demand, and norms of society all create a different context, which may affect selection criteria differently. To this end, the present study was intended to: 1) Identify and categorize factors affecting translators' book selection in the context of Iran; 2) Develop a scale to measure Iranian translator's book selection criteria; 3) Establish the psychometric properties of the TBSC scale as a reliable and valid instrument to assess influential factors affecting a translator's book selection behavior.

Book selection for translation and instrument validation
The following section provides an overview of relevant studies done so far. The first part deals with studies that report translation-related scales validation in Iran and the second covers research that examines book selection for translation.

Validating translation-related instruments
This section reviews the existing body of literature in the Iranian context that employs a factorial analysis approach to validate translation-related concept constructs. Several studies in various areas of audiovisual translation (AVT) have devised, tested, and validated an instrument for their purpose. The investigations conducted by Khoshsaligheh et al. are among the pioneering studies on AVT and literary translation in Iran (2014Iran ( , 2015Iran ( , 2017Iran ( , 2018. The perception of dubbing quality among Iranian dubbing viewers was investigated using a mixed-methods approach by Ameri, Khoshsaligheh, and Khazaee Farid (2018). Because there was no existing questionnaire that fit the purpose of their inquiry, they devised a 40-item questionnaire on a five-point Lickert scale to collect data. The initial item pool for the questionnaire was created via focus group interviews. To determine the validity of the findings, an exploratory factor analysis was performed. The participants were Iranian lay viewers who were asked to rank the dubbing quality standards along with their preferences for watching dubbed films. According to their research, the quality standards of dubbing reception factors can be divided into six categories: technicality (e.g., natural use of language in dubbing), agents (crediting the names of contributors from translators to dubbing actors), faithfulness to the stylistic features of the original content, content and visual censorship, domestication, and preferences for watching different dubbed content. Bijani, Khoshsaligheh, and Hashemi took another step in recognizing the Iranian expectation norm, this time in literary translation (2014). Using a mixed-methods approach, they designed and validated a questionnaire to measure the expectations of 424 university student readers in terms of acceptable renderings of foreign literature. The expectancy norms were categorized using exploratory factor analysis into six categories: visualization, source cultural items, target culture, authorial loyalty, target text preface, and target text language.
In another study, Khoshsaligheh, Pishghadam, Rahmani, and Ameri (2018) examined the relevance of emotioncy in Persian dubbing preferences. The concept of emotioncy deals with sense-induced emotions which might alter individuals' perception. To achieve this goal, they designed two self-report scales. The emotioncy scale, which ranges from 0 to 50, was created and used to assess the participants' emotions. The second scale was the dubbing lexical preference scale, which was developed to examine participants' preference in Persian dubbing. The findings of the study showed that emotioncy had a significant positive effect on the preference of the Iranian audience's choice for the lexical words included in the dubbed version.
In an attempt to explore violations of research ethics in Iran, Khoshsaligheh, Mehdizadkhani, and Keyvan (2017), constructed and validated an original questionnaire instrument. Their main goal was to find out how Iranian master's students of translation perceived and were aware of different types of research ethics violations. The validity and reliability of the selfdesigned instrument were determined using exploratory factor analysis and scale reliability analysis. They classified typologies of research ethics violation into five themes: no acknowledgment, excessive overuse, misreferencing, fraudulence, and duplicate publication. The data demonstrated that participants did not have an entirely accurate perception of research ethics violations.
Through an exploratory factor analysis approach, Salemi, Khoshsaligheh and Hashemi (2015) constructed a questionnaire to examine the cultural orientations of Iranian English translation students. They reported national identity, cultural heritage, local traditionalism, collectivism, and social attachment as the revealed categories of their participants' cultural orientation. Additionally, no significant relationship was found between participants' cultural orientation, their gender, and translation quality.
The purpose of the present investigation is to discover the underlying factors that influence the book selection criteria of Iranian professional translators. As little effort has been made to identify translators' book selection decisions in the Iranian context, the present study aims to contribute to addressing this gap. Furthermore, the study involved designing, validating and implementing a survey instrument that could be employed in similar studies in Iran.

Book selection for translation
To the best of the researchers' knowledge, no study has been conducted so far to examine the issue of book selection for translation from the perspective of translators. A few studies, however, consider the choice of books for translation from different aspects.
For example, a study by Abanomey, Al-Jabali, and Galal (2015) examined the ideal definition of a quality book (books considered by the academic staff of King Saud University to be suitable for translation). They also looked for variables that influenced selection criteria. The study revealed that participants disagreed on the ideal operational definition of 'quality book' and "the definition provided may reflect the influence of the respondents' majors, interests, attitudes, experiences or level of education" (2015, p. 46). Moreover, quality book selection criteria were identified in a top-down ranking: author and publication, book subject, book freshness, and title and content. Furthermore, the results suggest that the variables gender, book translation experience, and attendance at translation-related conferences significantly influenced participants' criteria selection. In contrast, the variables qualification and revising translated books did not show statistical significance.
In another study, Bieracka-Licznar and Paprocka (2016) identify and analyze the criteria and mechanisms for selecting books for translation from the publisher's point of view. The study found that "in terms of the participants in the literary translation event, it is quite surprising that two major actors -the translator and the literary agent -are largely absent from the process of selecting books for translation" (Biernacka-Licznar & Paprocka, 2016, p. 191). However, only the publishers make decisions at this stage. The study reported on three factors that influenced publishers' choice of French literature to translate into Polish: educational background, appreciation of French literature for children, and availability of funding. In Bourdieu's terms, the interplay of these three factors revealed that, for Polish publishers, "symbolic rather than economic capital" is important, as "their actions are driven by a book's beauty and refinement and not by ready profit. They give precedence to the quality of literature they release, for example through selection based on the industry's awards, rather than to revenue" (Biernacka-Licznar & Paprocka, 2016, p. 192).

Research design
A mixed-methods approach was adopted to develop an instrument to identify the Iranian translators' book selection criteria. The qualitative approach (the first phase) was conducted to generate an item pool through focus groups and literature review. During the second phase of the study, a pilot version of the questionnaire was developed, followed by validating the Translator's Book Selection Criteria (TBSC) questionnaire.

Participants
The main objective of this study was to identify the factors influencing translators' book selection in Iran. Any Persian native-language translator who has translated and published at least one book on the Iran publishing market, from or to any other language, and in any subject area may be part of the population of this study. The participants of the qualitative and quantitative phases were selected through snowball sampling. The participants voluntarily took part in the study. Moreover, to reach higher participation rates (Wright, 2005) and for the sake of time (McDonald & Adam, 2003), the questionnaire was designed and shared through a web-based platform.
For the qualitative section of the study, the researchers identified two translators who translated and published books in Iran. They were then asked to introduce other participants who had the same characteristics. This snowball sampling led to the identification of 16 other participants who voluntarily took part in the study. It should be noted that these 18 participants were not included in the second phase of the study.
To find the ideal participants for the quantitative part of the study, the researchers identified several well-known translators by examining newly translated books on the market. The researchers used Instagram and WhatsApp to communicate with the translators and provide information on the purpose of the study. The link to the questionnaire was sent to the translators who were interested in participating. They were also asked to forward the link to the other translators they knew. The data collection process lasted two months and resulted in the voluntary participation of 251 translators.

Item generation
The first step in developing a questionnaire is selecting concepts, i.e., variables, to be included in the study. The next task is devising items for each determined content area. As posited by Dörnyei and Taguchi (2010, p. 39), "question design is not a 100% scientific activity". However, to reduce the subjective nature of item creation, the researchers drew on three different sources in addition to their creativity to generate the item pool.
The researchers held three series of focus group discussions in Persian. The 18 translators were asked to name the factors affecting their book selection for translation, alongside the reasons behind each factor. Their arguments were audio recorded by cell phone to be transcribed for later processing. The gathered qualitative data were analyzed using the "three-level coding system, consisting of open coding, axial (or theoretical) coding, and selective coding" (Saldanha & O'Brien, 2013, p. 191). The analysis of the data revealed 27 factors affecting the participants' book selection for translation. As one of the sources, these focus-group-driven items were used to generate the item pool (see Table 1).

Book content
The original content must be of interest to the translator, be consistent with his/her knowledge & society's norms.

Reviews
Book reviews provide the translator with: -a summary of the content a critical assessment of the content information on whether readers would appreciate the book Awards Usually, award-winning books are better received by the audience.
High circulation High circulation in SL may indirectly be related to the books' popularity and reception in the original language.
Recommendation by social cataloging websites (e.g., Goodreads) It is a fast & safe way to find the best-selling books in the original language.
Recent publication It is prudent to pick newly published books because: -The content is novel, -It is less likely to be already translated Consultation with expert Due to their expertise, they know their field's gaps and the need for translation. This, in turn, may guarantee the translated book's sale.

Book size
Printing costs in Iran could be an obstacle. Hence, high volume books become costly. This may, in turn, reduce readers' purchasing power. Payment Even if I am not interested in the book's subject, a reasonable payment can convince me to translate it.

Gender-based selection
Some translators may show an inclination or prejudice against one gender, which may affect their choice.

Not being previously translated books
It is a risky decision to pick books that have already been translated. Because: -The content is not new for the reader anymore; - The translator may not be able to compete with the earlier translator, especially if he/she is a famous one.

-
Hence, the new translation may easily be ignored even without being read and compared with the previous ones.
Author-based selection It is helpful to translate the works of a particular author because: -You are familiar with his/her writing style, mindset, etc.

-
The author's name represents the quality of the book.

-
Translating the works of successful, well-known authors could bring parallel success & reputation for the translator. Dynamism The book selection for translation is not a static decision, even for a particular translator, because: -The translation is affected by society's taste & demand, which may change from time to time. -Some factors such as payment or publishing guarantee can make you change your fixed criteria. Commission by educational institutions such as universities These kinds of translations have the highest chance of being published.

Title
The title is what translators see first. Interesting titles capture translators' attention and help decide which book to select.

Content
The book's content must be of interest to the translator and under the norms of society.
Possibility of publishing People need to remember your name as a translator. This can be achieved by more and more publications. Otherwise, the translator becomes invisible in his/her profession.

Private publishers
They impose fewer restrictions on translators, provide better conditions for publishing, and offer a higher volume printing.
Public publishers They are more authentic, famous, and prestigious compared to private publishers.

Author reputation
The author's name is a good guide for both translators and readers in choosing the book.

Native written books
Translating from an intermediate language may affect the final product in terms of quality.

Translations into different languages
Usually, books translated into different languages are among the best, which worth translating into another new language.

Fixed criteria in book selection
Recognition of society's demand, do's, and don'ts of translation for a particular context, situation, and condition provide translators with a fixed formula of selection behavior.

Translators' freedom in selection
Although different factors influence translators' selection criteria, the translator makes the final decision freely.

Individual-driven book selection
Despite various influential factors, some translators still decide to work on their priorities.
Translators' inclinations Translators' habitus would determine their inclination toward unique books, authors, or subjects.

Translators' interests
Interest in the topic is the first and foremost factor in book selection for translation, which could also affect the quality of translation. Otherwise, the act of translation would become tedious.
Non-application of a theoretical framework in item generation would, according to Dörnyei and Taguchi (2010, p. 23), result in "an infinite number of items, all seemingly targeting important" aspects of the concept under study. Not to get caught in this trap, the researchers conducted a review of the related literature and identified several possible factors that could hinder or encourage translation book selection. The items developed based on the literature, the second source of item generation, are summarised in Table 2. Society's norm "Every phase of translation, from the selection of texts, to the adoption of an overall cultural stance and the specific strategies, is constrained by norms." (Xianbin, 2007, p. 24) Translator criticism "Regardless of the actual power of translators, text selection has often been an essential criterion of translator criticism." (Xianbin, 2007, p. 26) Translator' agency The way translators exercise their agency can affect the final translation product. (Pym, 1998) Ideology Translators themselves influence the ideology of translated texts i.e., "conforming to or rebelling against the dominant ideology." (Munday, 2016, p. 199) Original bestsellers A foreign best-seller would probably lead to higher economic capital. (Alkhamis, 2012) Society's demand Society's demands are so important for translators that they should be included in translation pedagogy. (Cui & Zhao, 2015) Borrowed questions were the third source used in item design in the present study. As mentioned earlier, there is no pre-constructed questionnaire on translators' book selection criteria in Iran. However, Haddadian-Moghaddam (2015), in his sociological study of translation in modern Iran, developed a 25item questionnaire to provide a sociological account of Iranian literary translators. Since 6 of the items in his research applied to the present study, they were included in the questionnaire. It must be acknowledged that the following three questions were borrowed from Haddadian-Moghaddam (2015): • The possibility of the work being published in Iran is important to me.
• As a translator, I tend to share the enjoyment of reading the work with others through translation. • I prefer to select original books for translation based on my knowledge of the authors and their works.
Besides, items 7, 8, and 9 of the second part of the questionnaire, i.e., the demographic section, were also borrowed from Haddadian-Moghaddam (2015). Finally, the results of the focus group discussions, the study of related literature, and the application of borrowed questions allowed us to identify six themes related to translators' book selection criteria in Iran, namely sociocultural considerations, external motivating factors, success guarantee factors, translator's authority, internal motivating factors, and publisher's authority. The researchers developed an item pool with 40 items based on the findings of the qualitative phase of the study.

Questionnaire construction
The issue of appropriate length is the initial consideration in developing questionnaires. The estimated time for the present questionnaire is about 20 minutes, which is the reasonable completion time for web-based questionnaires (Umbach, 2004). The initial questionnaire included 54 items and was divided into two parts. Part I contains 40 closed-ended statement-type content questions measured by six-point Likert scales ranging from strongly disagree (1) to strongly agree (6). The researchers decided not to include a middle point, i.e., a five-point Likert scale, due to the tendencies of Asian respondents to use the middle category while answering Likert-type questionnaires (Chen, Lee, & Stevenson, 1995). The demographic characteristics of the participants are dealt with in part II, which covers 14 closed-ended question-type items.
The selection of an appropriate informative title is another vital feature of well-designed questionnaires that helps respondents "to identify the domain of the investigation … and to activate relevant background knowledge and content expectations" (Dörnyei & Taguchi, 2010, p. 18). To satisfy these demands, Translators' Book Selection Criteria was chosen as the title of the questionnaire. In other words, the present researcher-made questionnaire intended to explore the criteria of book selection for translation among Iranian translators.
After designing the first draft of the questionnaire, content analysis was conducted at two different stages. During the first stage, to ensure face and content validity of the constructed questionnaire, three university colleagues in the field of Translation Studies were asked to answer the questionnaire in the researchers' presence. They provided feedback about the items regarding wording and clarity, deletion of unnecessary items, and inclusion of missing ones. This initial piloting led to the inclusion of a new item that considers book selection an entirely individual decision. Moreover, modifications were made to the wording of the three questions borrowed from Haddadian-Moghaddam (2015). In other words, to have a consistent writing style, e.g., to match the style of the borrowed items with those proposed by the researchers, the three colleagues suggested rewriting the items as follows: • I prefer to select books for translation with a higher possibility of publication in Iran. • I prefer to share the joy of reading the work with others through translation.
• I prefer to select original books for translation based on my knowledge of the authors and their works. Additionally, other modifications were also suggested by the three colleagues, which are presented in Table 3. I pick up an original book that has not previously been translated into the target language.
Deletion of universal 'never' to make the question clearer.
The original author's gender is influential in choosing it for translation.
When selecting a book for translation, I am sensitive to the author's gender.
To make the item clearer and more natural.
At the second stage of the content analysis, the researchers again sent the questionnaire link modified in the previous step to 10 experts in the field. These experts were university professors with doctoral qualifications who specialized in Translation Studies. They evaluated the level of relevance of each item for its corresponding construct on a 4-point scale (1 = not relevant, 2 = needs modification, 3 = relevant but needs radical modification, 4 = highly relevant). Moreover, the level of essentiality of each item for its corresponding construct was examined on a 3-point scale (1 = essential, 2 = useful but not essential, 3 = not essential). A modification was made to the items that were judged to need modification. However, none of the items was removed at this stage (Section 4.2.4.).
After finishing the content analysis of the questionnaire, the researchers used Porsline (https://survey.porsline.ir/) to create the web-based questionnaire. Following Couper, Traugott, and Lamias (2001), a progress indicator was specified to show participants how much is left to complete. Moreover, to prevent respondents from skipping a response, the questionnaire was programmed not to deliver an item before the previous one has been answered. Hence, the 40-item web-based questionnaire was ready for distribution.

Pilot study
Upon preparing the first version of the questionnaire, it was time to pilot the questionnaire with participants similar to the target population. At this stage, five translators who translated and published at least one book were identified and completed the web-based questionnaire voluntarily. They all had Persian as their native language and translated into other languages, including English, French, and Italian. Translation directionality was not considered. They were asked to forward the questionnaire link to other translators who met the criteria.
The snowball sampling, which resulted in a pilot sample of 50 translators, was used for data gathering.
The reliability of the scale was estimated via Cronbach's alpha coefficient (Appendix A). The result was .635, which, according to some researchers (e.g., Nunally & Bernstein,1994;Taber, 2018), is both satisfactory and sufficient, particularly in exploratory analyses (e.g., pilot testing).

Validation of the TBSC questionnaire 5.2.1. Item analysis
To establish the psychometric properties of the Translators' Book Selection Criteria (TBSC) scale, the following analyses were carried out: (a) Descriptive statistics for summarizing group characteristics and demographic distributions, (b) Cronbach's alpha coefficient for reliability measurement, (c) Lawshe's content validity analyses, (d) Principal component analysis with Promax for validation and factor extraction, and (e) Monte Carlo simulation for parallel analysis.

Reliability assessment
The reliability of the scale was estimated via Cronbach's alpha coefficient (Appendix B). Cronbach's alpha is the most common test to evaluate instrument reliability in L2 research, particularly with composite variables (Plonsky & Derrick, 2016). The result was .745, which, according to George and Mallery (2010), is satisfactory and acceptable.

Content validation
Content validity was established prior to the main analysis to determine the essentiality and relevance of the items. It was carried out by distributing the TBSC to 10 experts in the field and running Lawshe's (1975) CVI (content validity index) and CVR (content validity ratio) analyses.

Validation
Principal component analysis (PCA) is a form of exploratory factor analysis to explore the underlying structure of the intended variables (here the items of the TBSC checklist). It is a sophisticated tool to refine (and reduce) items of newly developed checklists to form a more manageable number of components or subscales; it removes the items that weaken the construct validity of a checklist (Loewen & Gonulal, 2015). In this study, two forms of PCA were run, one based on the eigenvalues and the second one using a fixed number of factors (George & Mallery, 2010).

Data clean-up 5.2.5.1.1. Missing values
The Likert-type checklist items (1-40) were scanned for unusual response patterns and missing values. The items were found to be answered acceptably since there was no instance of identical response patterns to positively and negatively keyed items. However, there were six cases of nonresponses (2.98 %) in the raw data.
Running factor analysis with missing data is problematic; yet, multiple imputation or deletion (listwise or pairwise) methods are not recommended here for the following reasons: (a) The number of the nonresponses was not large enough, thus violating the main assumption of multiple imputations (< 5%); (b) The missing values were of MCAR (missing completely at random) type, in which case respondents with missing data were not systematically different from those with complete response patterns. Thus, to avoid reducing sample size and minimize the possibility of data bias, the mean imputation method (Enders, 2010) was adopted for treating the nonresponses. It was done by replacing the missing values with the mean for each item (i.e., average score of the available cases at the item level).

Univariate outlier detection
The next step was to check the data for statistical outliers. After computing the z-scores, they were scanned for the occurrence of any extreme, probable, and potential outliers. According to Field (2013), in normal distributions, nearly 95% of the scores should fall in the acceptable range, and the remaining 5% (or less) can be either potential or probable outliers; no extreme outliers should occur. To detect the outliers, Field's (2013) syntax was run in SPSS. The results revealed that the first two assumptions were accurately met since more than 95% of all the cases fell within the normal range (z < 1.9), and 3% were potential or probable outliers (1.9 < z < 3.2). Yet, in eleven items (i.e., Items 9,13,17,19,26,27,28,31,33,35,37) extreme scores were detected. To minimize the possibility of skewness and enhance the accuracy of the results, the winsorizing approach was adopted by transforming the outliers to a high value that was not an outlier (Field, 2013).

Preliminary analyses
Before the main analysis, the three major assumptions of PCA, namely linearity, skewness, and kurtosis, were examined. Linearity was assessed by doing a spot check of the combinations of variables. Since no curvilinear relationship was observed, it was confirmed that the relationship between the variables was linear. The skewness and kurtosis z-scores were also investigated and ensured to fall within the acceptable range of ± 3.29.

Initial PCA and parallel analysis
Once PCA was run on the 40 items of TBSC, the data was found factorable on the following grounds: (a) Kaiser-Meyer-Olkin value was .813, which fell within the acceptable range (.7 ≤ KMO ≤ 1), suggesting that the sampling was adequate (Appendix C). (b) Bartlett's test of sphericity was statistically significant (X 2 = 3154.704, df = 780, p < .001) (Appendix C). (c) Correlations between variables were neither too high nor too low (correlation matrix here displayed a reasonable number of coefficients greater than .3).
(d) Variables had acceptable communalities (well above .4), thus imposing no item deletion at this stage (Appendix D).
When the initial PCA (based on the eigenvalues) with oblique rotation (Promax) was conducted (k = 40; no item deletion had been done yet), it produced 11 components with eigenvalues above Kaiser's criterion of 1.0 (Appendix E). The 11 variables accounted for 59.89% of the total variance. Yet, an inspection of the scree plot suggested a clear break between the fourth and fifth components, recommending the retention of the first four components; moreover, there is a moderate break after the eighth component, implying that seven-or eight-factor solutions were also plausible (Appendix F).
Finally, Horn's (1965) parallel analysis was run using the Monte Carlo simulation program. In a parallel analysis, the obtained eigenvalues are compared with those computed in PCA. Those eigenvalues in PCA that exceed the corresponding values in the parallel analysis represent the suggested number of components. The results indicated that four components should be retained (Appendix G).

Follow-up PCA and factor extraction
When the follow-up PCA was performed (with a fixed number of components), the percentage of variance explained by the four extracted components dropped to 38.26%. Moreover, 21 items were found to have low communalities (< .4), and 13 items loaded on more than one component, suggesting the removal of more than half of the items (Appendices H-J). Thus, it seemed that the fourcomponent solution underestimated the variance and was thus inappropriate.
Another PCA with a fixed number of components was carried out, forcing an eight-factor solution, as hinted earlier in the scree plot (Appendices K-O). The results displayed an enhancement in the percentage of the variance explained and item maintenance: The eight components explained 51.83% of the variance (Table 4), and only four items had communalities lower than the criterion (i.e., Items 9, 16, 25, and 39).
It is recommended that the retained factors should have at least three items with a loading greater than .3 (Field, 2013;Raubenheimer, 2004;Tabachnick & Fidell, 2014). After Items 9, 16, 25, and 39 (due to low communalities) were removed, Component 6 with two items (Items 7 and 15) and Component 4 with one item (Item 4) remained, which needed to be either revised or removed. Items 4 and 7 were removed since (a) the removal of Item 4 contributed to the scale reliability, and (b) Item 7 did not load on any other component (it did not meet the CVI and CVR criteria in Section 4.2.4. either). Item 15 cross-loaded on two components with loadings greater than .3, and its removal did not result in substantially higher reliability; thus, it was retained and assigned to Component 6.
Overall, the final model contained 34 items loading on six components (Appendices P-R). It accounted for 49.83% of the variance (Table 4), with Component 1 explaining 17.95% (Items 31,28,5,27,35,19,17,13,3,24,26,12), Component 2 explaining 13.54% (Items 33,36,37,20,32,29,6), Component 3 explaining 6.54% (Items 18,11,22,14,38,10), Component 4 accounting for 4.34% (Items 30, 23, 1), Component 5 explaining 3.77% (Items 34, 2, 40), and Component 6 explaining 3.67% (Items 21, 8, 15) ( Table 5). The reliability of the six-factor model was re-examined ( Table 6). The corrected TBSC showed a reliability index of .748, which is slightly higher than the reliability obtained before item deletion (i.e., α = .745).  After reviewing the items loading on each component, the subscales were labeled as follows: Component 1 was entitled General Considerations, Component 2 external motivating factors, Component 3 success guarantee factors, Component 4 translator's authority/agency, Component 5 internal motivating factors, and component 6 publisher's authority/agency. (Table 7) Table 7: The subscales of TBSC as labelled Subscale 1: General Considerations 1-I prefer to select books for translation with a higher possibility of publication. 2-I think original book selection for translation is affected by society's demands. 3-I think society's norms govern original book selection. 4-I prefer to translate books that have received good reviews. 5-I prefer to share the joy of reading the work with others through translation. 6-I think the content of the book is crucial for me to choose for translation. 7-I prefer to translate original books written by non-native authors. 8-I prefer to choose original books for translation based on my inclinations. 9-I prefer to choose original best-sellers for translation. 10-I prefer my translation to be published by a famous publisher in Iran. 11-I prefer my areas of interest as a translator to be shown in my original book selection. 12-I prefer to work with public publishers. Subscale 2: External Motivating Factors 13-I prefer to translate award-winning books. 14-I think the most crucial factor is the payment suggested for that translation. 15-I prefer to choose source language high circulation books for translation. 16-I prefer to translate books made into movies. 17-I think the author's gender is a factor to be accentuated when selecting a book for translation. 18-I prefer to translate original books recommended by social cataloging websites such as Goodreads, etc. 19-I prefer to choose original books recommended by the target language publisher. Subscale 3: Success Guarantee Factors 20-I prefer to consult with experts in the field when choosing a book. 21-I prefer to translate an original book that has also been translated into other languages. 22-I prefer to translate original books issued by a famous publishing house in their home country. 23-I prefer to translate newly published books. 24-I think the title of the book helps select it for translation. 25-I prefer to translate original books commissioned by national institutions such as universities.

Discussion
The final version of the TBSC questionnaire included 34 close-ended questions loaded on six different components. The researchers named each subscale considering its constructing items. As the most comprehensive, the first subscale covers the various considerations that come to the translator's mind, from the social considerations to the qualities of the book that may affect its possibility of publication or higher rate of reception among the target audience.
Some individual concerns are also included in this subscale, for example, the translator's area of interest or sharing the joy of reading an original book, which could be powerful motivators for a translator to select a book for translation. A book being a best-seller or having received good reviews is also among the first factors that may be taken into account by a translator when considering a book for translation.
The items of the second subscale refer to external factors promoting the translator's selection. In other words, the translator considers the book itself and external motivating factors such as payment, recommendation, high circulation of the source language into account since these external factors are believed to contribute to the prominence and value of the translated book.
The third subscale covers factors that could guarantee the translator's success and achievement. In other words, translating a book with one of the features listed in subscale three could be a wise selection, since experts in all fields and educational institutions such as universities recognize the gaps in their field (Kafi, Khoshsaligheh & Hashemi, 2015). Therefore, when they commission a book for translation, the possibility of publishing and republishing is very high. This is also the case with newly published original books due to the freshness of the ideas. Besides, the book's title can play a guiding role for both the translator and the reader. Engaging eye-catcher titles might attract people's attention at first glance.
The next subscale included factors related to the translator's authority or agency. Mingjian and Yu (2003, p. 22) believed that the translator's agency is: manifested not only in the translator's comprehension, interpretation, and artistic re-presentation of the source texts, but also in the selection of source texts, the cultural motivations of translation, the adoption of strategies, and the manipulation in the prefaces of the expected functions of the translations in the target culture.
Every translator has specific objectives and interests in mind, both personal and collective. Some prefer to remain faithful to society and market norms, but some may prefer to challenge them. For this group of translators, the determining authority is the translators themselves. This is called freedom or individual decision in selecting a book for translation. Besides, translators may, for example, decide to translate the works of famous authors in search of a reputation in line with that of the source author.
Having general knowledge of the author, his/ her works, and writing styles could be personal incentives for some translators that might be helpful in book selection (Shaki & Khoshsaligheh, 2017). Feeling close to the book's content and seeing it in harmony with one's inner self or personal ideology can help the translator in the process of translating and selecting the book for translation (Khoshsaligheh, 2013). Apart from these reasons, the volume of the chosen book can also be another vital factor for the translator, considering various limitations such as time and scheduling constraints, for which the translator is the primary decision maker. These factors are included in subscale 5.
The last subscale concerns the publisher's authority or agency. Hermans (1999b, p. 74) argued that "translation decisions are neither fully predetermined nor totally idiosyncratic". Publishers, as we know, are a major part of the patronage system at work in the translation process, determining the choice of social factors and the kinds of translations they would like translators to produce. These criteria may vary from time to time, as publishers prefer to work in line with what society chooses to be present and compete on the market. They are also concerned with the afterlife of the translation; one of its embodiments is translation criticism, which can affect both the translator and the publisher, since "regardless of the actual power of translators, text selection has often been an important criterion of translator criticism" (Xianbin, 2007, p. 26).
After naming each subscale, the sequence of items is another crucial issue to consider since, as argued by Dörnyei and Taguchi, "the context of a question can have an impact on its interpretation and the response given to it" (2010, p. 46). Considering the 6 multi-item subscales of the questionnaire, it was decided to start with general consideration subscale as it concerns comprehensive issues relating to sociological, cultural, and personal aspects of book selection. Items supported by external motivating factors were placed in the second subscale, followed by items that could guarantee the translator's success as the third subscale. The fourth subscale was devoted to items relating to the translator's agency. Items concerning the issue of internal motivating factors were included in subscale five. The last subscale included items reflecting the publisher's authorization. Moreover, different items of each subscale were separated at the furthest possible distance.

Conclusion
To the best of the researchers' knowledge, the present study is the first to develop and validate an instrument to measure translators' book selection criteria in the context of Iran. The study identified 34 items that Iranian translators might take into account when selecting a book for translation. The items were categorized under six components: general considerations, external motivating factors, success guarantee factors, translator's authority/agency, internal motivating factors, and publisher's authority/agency. Furthermore, the study's findings showed that the TBSC is a reliable (i.e., α = .748) and valid instrument for measuring translators' book selection criteria. Although more research is needed to strengthen the future development of the TBSC, the researcher-constructed scale could be used as a starting point to gather data on the issue of selecting books for translation in Iran, since the socio-cultural aspects of book selection for translation may vary from one society to another. The findings in this investigation should be interpreted in the context of some limitations. First, we utilized a non-random sampling due to the challenges in recruitment of eligible participants. Further studies may be conducted using a wider range of participants. This would enhance the generalizability of the findings (Ferdowsi & Razmi, 2022). Second, other statistical techniques such as SEM analyses can be utilized to conduct CFA, provided that the number of participants is sufficient for statistical analyses. Moreover, the designed instrument in this study can be used in other contexts to ensure the external validity of the instrument. The present instrument could also be used as a guide to identifying translators' culture-specific criteria of book selection in different societies to construct a new instrument for different contexts. In other words, to adapt the scale for use in other communities and cultures, there may be a need for substantial modification, a complete rewriting of items, and the omission of some of the current items or inclusion of new ones. ******************************************************