Linguistic Politeness and its Relationship with Data Collection Preferences

1. Introduction

Data collection which aims at collecting information systematically on variables of interest in order to achieve integrity in research is under the influence of several factors such as sampling process, design of the study, social interaction between the researcher and the participants of the study, the context of the study, the field of study, the applied procedure, and researcher’s biases and perceptions (Sapsford & Jupp, 2006).  Moreover, it seems that data collection in research could also differ according to cultural features (Pishghadam, 2013).

One of the cultural factors deserving meticulous attention is politeness which is a fundamental part of culture shaping human behavior within a society; however, substantial amount of research indicated that the conceptions of politeness vary across cultures (Almursy& Wilson, 2001; Ide, 1989; Janney & Arndt, 1993; Lee-Wong, 2002; Matsumoto, 1989; Nwoye, 1989). Cultural differences and similarities in the endorsement of rules for politeness can be explained by features such as cultural individualism-collectivism, low-context-high-context, directness-indirectness, analytic reasoning-synthetic reasoning, etc.

The terms ‘individualism’ and ‘collectivism’ (Dumont, 1986; Hofstede, 2011; Lukes, 1973) have gained the highest level of popularity in cross-cultural psychology as these constructs have been successfully implemented for the description, illumination, and prediction of varieties in the realm of values, cultural patterns, attitudes, cognition, social systems, morality, self-concepts, and ideology (for an overview, see Hofstede, 1980, 2001; Kagitçibasi, 1997; Miller, Bersoff, & Harwood, 1990; Oyserman, Coon & Kemmelmeier, 2002; Triandis, 1995;Witkin& Berry, 1975).To illustrate, people in collectivistic cultures prioritize the ingroup’s goals over individuals’ goals, as opposed to people in individualistic cultures who emphasize individuals’ goals over the ingroup’s (Triandis, 1995). Furthermore, using low-context communication is common among people in individualistic cultures, while resorting to high-context communication is common among people in collectivistic cultures (Gudykunst & Ting-Toomey, 1988). Therefore, cultural individualism-collectivism is believed to have impacts on the degree of emphasis placed by politeness rules in a culture on high/low-context communication (Ogawa & Gudykunst, 2000) -e.g., being direct, assertive - or high-context communication- e.g., not being too verbal, refraining from direct messages and confrontations. 

Given that there is a correlation between the level of indirectness and politeness (e.g. Brown & Levinson, 1987; Leech, 1983) on the one hand, and the influence of cultural features on data collection (Pishghadam, 2013); the ways members of a culture perceive politeness rules could affect their data collection preferences. With that in mind, in order to examine the relationship between conceptions of linguistic politeness and data collection preferences, the present study tries to design and validate a scale of Data Collection Preferences (DCP) along with a scale of Conceptions of Linguistic Politeness (CLP).

2. Theoretical Framework

2.1 Research Methods

It is important to identify the research design of a study as it entails information about key features of the study, which can be different for qualitative, quantitative, and mixed methods. Crotty (1998) outlined four key aspects to consider in a research design: the epistemology informing the research, the philosophical stance (e.g., post-positivism, constructivism, pragmatism, advocacy/participatory; see Morgan, 2007) underlying the methodology in the study, the methodology itself, and the techniques and procedures applied in the research design for collecting data.

Although data collection process is commonly applied in research, it is implemented in different ways, and for different purposes. Well-chosen and well-implemented methods for data collection and analysis are of utmost importance for all types of research. Therefore, qualitative studies try to collect and analyze qualitative data; quantitative studies resort to collecting and analyzing quantitative data; and so on.

Based on Crotty’s (1998) categorization, the description of research designs can be as qualitative, quantitative, and mixed methods. In qualitative research methods, there is a focus on discovering and realizing the experiences, perspectives, and thoughts of participants, so qualitative research explores meaning, purpose, or reality (Hiatt, 1986). Qualitative research can be considered as a situated activity locating the observer in the world. It encompasses a set of interpretive, material practices such as “a series of representations, including field notes, interviews, conversations, photographs, recordings, and memos to the self” in order to make the world visible, thus, qualitative researchers study things in their natural settings with the aim of making sense of, or interpreting a phenomenon in terms of the meanings people attribute to it (Denzin & Lincoln, 2005, p.3).The fundamental feature of this inquiry is the existence of multiple ‘truths’ which are socially constructed (Lincoln & Guba, 1985). The nature of the qualitative research allows for a detailed exploration of a topic of interest in which information can be collected by a researcher through different data collection methods such as case studies, ethnographic work, interviews, protocol analyses, observations, narrations, and so on. The main premise in this approach is that a flexible and open research process and results driven inductively would be the outcome of the description of a series of the interactions among researchers and participants of the study in real settings (Harwell, 2013). Thus, qualitative research doesn’t generally set “replicability and generalizability” as its goals (Harwell, 2013, p. 149).

On the other hand, quantitative research methods strive to maximize objectivity, replicability, and generalizability of findings which could facilitate prediction. In this approach, the researcher is expected to set aside his/her experiences, perceptions, and biases for ensuring objectivity in performing the study and drawing the conclusions. The use of instruments such as tests, surveys, questionnaires, checklists, and so on are common in quantitative studies to collect data. Inherent to quantitative studies is testing statistical hypotheses related to the research questions of interest based on the probability theory (Harwell, 2013). Since general inferences about a given population would be achieved through inferences made from tests of statistical hypotheses, these types of inquiries are mostly deductive in nature (Harwell, 2013). Furthermore, the underlying assumption in quantitative methods is that there exists a single ‘truth’, independent of human perception (Lincoln & Guba, 1985, as cited in Harwell, 2013).

Along with the qualitative versus quantitative dichotomy, mixed methods seek to combine the qualitative and quantitative methods in order to make the most of their differences for dealing with a research question (Harwell, 2013). Mixed methods can be traced back to the multi-trait, multi-method approach of Campbell and Fiske (1959, as cited in Teddlie & Tashakkori, 2009), in spite of being regarded as a relatively new methodology whose basic philosophical, methodological, and practical standards have evolved since the early 1990s (Tashakkori, 2009, as cited in Harwell, 2013).The advantage of using mixed methods research over the sole use of qualitative or quantitative data lies in collecting multiple kinds of data with different strategies and methods with strengths (Johnson &Turner, 2003, as cited in Harwell, 2013). In other words, mixed methods research provides the “opportunity to compensate for inherent method weaknesses, capitalize on inherent method strengths, and offset inevitable method biases” (Greene, 2007, p. xiii).

2.2 Politeness

The enquiry into “what might be labelled linguistic ‘politeness’ entails the social dynamics of human interaction” (Kadar & Culpeper, 2010, p. 9).Scrutinizing the field of politeness studies in linguistics and particularly in pragmatics; we may identify two broad theoretical approaches to this concept. The first one can be called the ‘traditional’ view having the status of ‘classics’ in the field traces back to Gricean and speech-act theoretic view (Brown & Levinson, 1987; Lakoff, 1973; Leech, 1983). In this approach, there seems to be a shift from actual speakers to supposed model persons featured with rationality and face(Brown & Levinson, 1987; Lakoff, 1973; Leech, 1983). The underlying assumption in the traditional perspective is to view different cultures as internally homogeneous with regard to politeness realization and definition and as a result its measurement (Brown & Levinson, 1987; Lakoff, 1973; Leech, 1983).

Goffman (1955, 1967) was the pioneer in carrying out the first groundbreaking studies on interactional ‘facework’.  In Goffman’s (1967) term, ‘face’ is defined as “the positive social value a person effectively claimsfor himself [sic] by the line others assume he has taken during a particularcontact” and ‘facework’ is regarded as the communicative action orienting to the speaker’s and/or the addressee’s face (p.5). In the 1970s, the sociopragmatic study of ‘politeness’started most notably with Lakoff’s groundbreaking papers (see Lakoff 1973, 1977). In the1980’s, two seminal works were published, one by Brown and Levinson (1978, 1987) and the other by Leech (1983).Leech’s politeness  principle  tries to compensate for  the  missing  link  between  the  Gricean Cooperative  Principle  and  the  problem  of  how  to  relate  sense  to  force (Leech, 1983).

However, Brown and Levinson's theory of politeness is considered as the most influential in offering a framework for the study of linguistic politeness (Watts, Ide, & Ehlich, 1992).Their paradigm is known as the face-saving view which differs from that of Leech in the sense that  Leech's  approach  of  politeness  sets in  a  more  general  pragmatic  theory, whereas Brown  and  Levinson’s  proposal  is a  comprehensive  theory  in which specific politeness strategies are realized through linguistic devices (Fraser, 1990).The theory of politeness proposed by Brown and Levinson (1987) is derived from three basic notions:  “(a) the view of communication as a rational process, (b) Gricean Cooperative Principle and the conversational maxims, and (c) Goffman’s (1967) notion of face” (p. 61).

Despite its seminal contribution to the field of linguistic pragmatics, Brown and Levinson’s (1987) politeness model has been subjected to criticism by many researchers mainly on the grounds that the model is Anglo-biased, oriented towards individualistic, egalitarian cultures in the west rather than group-based, collective, hierarchical cultures of some eastern communities (Watts et al., 1992).

Scollon and Scollon (1995) were among the scholars who took a cultural stance about politeness by stressing the assessment of the appropriate level of face in communication. They note that politeness is under the influence of factors such as “power, distance and the weight of the imposition” (p. 52).  Scollon and Scollon (1995, pp. 54-55) postulate that based on the power difference (+p,-p) and the distance (+D,-D) between the participants, we have three types of politeness system:

1. “Deference politeness system (-P, +D)”:Although the participants in this type of politeness system are considered to be of the same rank, they are at a distance. This system of relationship is characterized by being symmetrical (-P) at the equal social level and distant (+D) by applying independence strategy. For instance, two professors from two different countries meeting at a conference adhere to this system.

2. “Solidarity politeness system (-P, -D)”:There is considerable use of involvement strategies here, as the participants feel lack of power difference (-P) and distance (-D) between themselves.  The Solidarity politeness system is characterized by being symmetrical (-P) through having the same social level and close (-D) by applying involvement strategies. Two close friends are an example.

3. “Hierarchical politeness system (+P, +/-D)”:Hierarchical politeness system is mainly observed when there is difference in social positions of the participants in the sense that one is in the super-ordinate position and the other in subordinate level. This system is characterized by being asymmetrical regardless of the distance seen between the participants. If the person is in a higher position, he/she applies involvement strategies while the person in the lower level implements independence strategy. This system is commonly seen in business, official and governmental relationships.

Since 2001, a new ‘school’ has been established within politeness research as the ‘postmodern’ or ‘discursive’ approach (e.g., Eelen, 2001; Mills, 2003; Watts, 2003)which has questioned many of the proposed concepts in the traditional approach to politeness (Kadar & Culpeper, 2010). The post-modern approach seeks to tackle the challenge introduced by increasing empirical evidence for the traditional view. Influenced by social theory, the post-modern view addresses the nature of politeness norms across cultures, and, basically, within cultures through stressing the role of the addressee in negotiating politeness at the micro-level jointly with the speaker (Eelen, 2001; Mills, 2003; Watts, 2003).

The post-modern theories’ focus on first-order politeness - people’s common definitions of, and meta-linguistic judgments about- as well as the importance of situated evaluation for politeness- incorporating Bourdieu’s notion of habitus- inspired them both to reject the Gricean and speech-act theories (Eelen, 2001; Mills, 2003; Watts, 2003). The implication of adopting 'post-modern theories’ is the fact that politeness is not always considered as a 'good' matter by participants (Watts, 2003); i.e. politeness itself can be assessed, and it can be regarded either a positive or a negative value denoting the  attempt to manipulate a situation for one’s own benefit (Terkourafi, 2005).

In short, the essence of postmodern perspective of politeness is a departure from following an “a priori predictive theory of politeness or a post-facto descriptive theory of politeness” (Watts, 2003, p. 142),as politeness is believed to be situation-specific. Furthermore, in this theory, it is believed that politeness and impoliteness should be considered as part of a continuum rather than polar opposites (Eelen, 2001; Mills, 2003; Watts, 2003); thus, a comprehensive theory on politeness should deal with covering both impolite and polite phenomena within its framework.

2.3 Setting the Scene: Iranian Society

Living in an Eastern society, Iranians tend to enjoy and emphasize less individuality than Westerners, who favor individualism or self-identity over collectivity or group-identity. To illustrate, Kinnison (2001) contends that people in Eastern cultures consider themselves as part of networks (group-identity) rather than being merely individuals (self-identity). The main aspects of social relationships in the Iranian culture are group consciousness and interdependence (see Koutlaki, 2002). This cultural preference for collectivity in Iran can trace back to the existence of extended families until recently and clans or tribal units until last century, as well as Iranians’ rich social traditions and customs.

Thus, Iranian culture with rating 41 is closer to the collectivist than the individualist end of Hofstede’s continuum of ‘individuality vs. collectivity scale’ (Hofstede, 2001).The concept of individualism is alien to Iranian culture (Eslamirasekh, 1993), due to the fact that in the Persian language, the term ‘individualism’ bears negative connotations and nearly carries the sense of selfishness and self- centeredness as opposed to ‘collectivism’ emphasizing  the interdependence of all members of society as well as the well-being of the group. Furthermore, in line with Triandis’ (1995) typology of individualism and collectivism on the basis of horizontal and vertical social relationships, It seems that adhering to horizontal collectivism in which interdependence, sociability, and common goals are highlighted (Singelis, Tdandis, Bhawuk, & Gelfand, 1995) is more common among Iranians rather than vertical collectivism which stresses hierarchical relationship (Chen, Meindl, & Hunt, 1997).

Based on Nanbakhsh’s (2011) study, Iran is mainly a positive-politeness-oriented society which differs in its attitude to politeness from noticeably negative-politeness oriented societies. Similarly, Eslamirasekh (1993) postulates that “The use of positive politeness strategies in Persian stems from the value of group orientedness in Iranian culture” (p. 97). Consequently, in a society such as Iran where group values and collectivity are dominant, it is essential to improve, empower and sustain social ties and connections with other members of the society.

3. Purpose of the Study

As it seems that cultural features such as politeness conceptions could affect data collection in research (Pishghadam, 2013), the present study is aimed at examining the relationship between conceptions of linguistic politeness and data collection preferences among Iranians. In so doing, the study seeks to design and validate two scales: a scale of data collection preferences along with a scale of conceptions of linguistic politeness. 

The politeness scale was designed based on Scollon and Scollon’s (1995) theory of deference, solidarity, and hierarchical politeness systems. In order to check the construct validity of politeness scale as well as research scale, Confirmatory factor analysis (CFA) was utilized and their reliability was checked using Cronbach's alpha. Finally, the relations between the CLP and the DCP scales are examined using Pearson product-moment correlation to shed more light on their nature.

4. Method

4.1 Participants

First for the purpose of designing and validating an emic scale of Iranians’ conceptions of linguistic politeness, a total of 502individuals participated in the present study. They included 262 females, and 240 males. Their ages varied from 15 to 71. In line with the purpose of the study, the participants were chosen from different ages, educational levels, and occupations. Among this population, 199 individuals- 127 females, and 72 males- participated in the second phase of the study for designing and validating a scale of data collection preferences. They were either graduates or still students at M.A., M.S. or Ph.D. levels at different majors at different universities of Mashhad, Iran- Ferdowsi University of Mashhad, Imam Reza International University, and Islamic Azad University. The assumption underlying the selection of such participants was that, since in Iran, people practically become more familiar with research practice and its goals in post-graduate studies, they have a better understanding of research data collection methods at these levels. The participants were selected based on their willingness to participate.

4.2 Instrumentation

4.2.1Conceptionsof Linguistic Politeness (CLP) Scale

The authors developed and designed a scale (in Persian) based on Scollon and Scollon’s (1995) theory of deference, solidarity, and hierarchical politeness systems. There were 22 items in the initial model, Hierarchical Politeness System (HPS) = 8 items, Deference Politeness System (DPS) =4 items, and Solidarity Politeness System (SPS) = 10 items, as indicators of different types of politeness systems proposed by Scollon and Scollon (1995). HPS implies the situation in which there is an asymmetrical relationship (+P) between the participants, while they may or may not be at a distance (­­-/+D). DPS denotes the relationship between the participants which is symmetrical (-P) and distant (+D), and finally SPS indicates the situation in which the participants have a symmetrical (-P) as well as close (-D) relationship.

Writing the items involved three steps. First, for assuring the content validity of the scale, a comprehensive review of the literature was done by the authors and the three types of politeness systems proposed by Scollon and Scollon (1995) were specified. According to Scollon and Scollon (1995) based on the power difference (+P,-P) and the distance (+D,-D) between the participants, there can be three types of politeness systems: deference politeness system (-P, +D), solidarity politeness system (-P, -D), and hierarchical politeness system (+P, +/-D).Based on these politeness systems, items addressing deference, solidarity, and hierarchical politeness systems were written. Then, these features were operationalized and modified for Iranian society. Items refer to different situations in which a type of request or action was made. Participants responded to the items on a 5 point Likert-type scale ranging from 1 (most impolite) to 5(most polite). Afterwards, the CLP scale was piloted with 10 individuals to check its reliability as well as the appropriateness of the content. Having completed the questionnaire, the participants were asked to talk about the content of the scale, and whether it is comprehensible for them. Some modifications and rewordings were done in the items according to the participants’ viewpoints. Then, the scale was administrated to another group of (n = 38) participants to check its internal consistency. Cronbach’s alpha achieved for this sample was .89 which assured the researchers to proceed with data gathering to assess Iranians’ conceptions of linguistic politeness. The reliability and validity of the scale for the main study are stated in the results section.

4.2.2 Data Collection Preferences (DCP) Scale

The authors developed and designed a scale (in Persian) based ondifferent data collection methods in research. Accordingly, 19 items entailing different research data collection methods were written. In writing the items, three steps were followed. First, having reviewed the related literature, for assuring the content validity of the scale, the authors tried to specify common research data collection methods applied in Iran.  Moreover, the authors divided the items into explicit and implicit. The explicit items entail those in which there were direct verbal messages such as closed questionnaire, structured interview, protocol analysis while doing an activity, direct research questions, direct self-report, and direct observation. The implicit items included those in which there were indirect verbal messages such as open questionnaire, unstructured interview, and protocol analysis after doing an activity, indirect research questions, metaphor, narration, indirect self-report, and indirect observation. The assumption underlying such division was that since Iran is considered as a collectivistic country (see Eslamirasekh, 1993; Hofstede, 2001; Nanbakhsh, 2011) with the features of a high-context society, there is a focus on indirect verbal messages (Hall, 1976) observing politeness rules. Having considered different data collection methods in Iran and the explicit and implicit dichotomy, items reflecting these features were written.

Afterwards, the items were operationalized and modified for the research settings.  The items refer to the participant’s preference for attending different research data collection methods. Participants responded to the items on a 5 point Likert-type scale ranging from 1 (completely disagree) to 5 (completely agree). Then, 10 individuals took the DCP scale at the pilot phase of the study through which the reliability of the scale along with the suitability of the content were checked. Afterwards, some modifications and rewordings were performed in the items based on the participants’ opinions about the comprehensibility of the items. Subsequently, in order to check the internal consistency of the scale, it was administrated to a group of 34 participants. Having calculated the internal consistency of the scale through Cronbach’s alpha which was .84 for this sample, the researchers were ascertained to continue data gathering to measure Iranians’ data collection preferences. The results section has dealt with the reliability and validity of the scale for the main study.

4.3. Procedure

The CLP scale was given to different individuals with different ages as well as different educational and occupational backgrounds. Among these participants those who graduated or were studying at post-graduate levels were also given the DCP scale. Participants were informed that it was voluntary to fill the scales. Since Persian was the first language of all the participants, both scales were written in Persian to ascertain that they can understand and the content of the scales correctly and to increase their response rate. It took about 25 minutes to complete the two scales.

First, Amos 20 was utilized to check the construct validity of the scales using Confirmatory Factor Analysis (CFA) method. CFA takes a confirmatory hypothesis-testing approach to the data. Then, the fit of the model is assessed with different goodness of fit indices. In the present study, χ2/df, GFI, CFI, and RMSEA were used.

Then, in order to assess the relationship between the CLP and the DCP scales, Pearson product-moment correlation formula was used.

5. Results

This study was conducted to construct and validate a scale of conceptions of linguistic politeness as well as scale of data collection preferences, and to examine the relationship between the two devised scales.

5.1Validating the CLP Scale

In order to assess the validity of the scale, CFA was used. Different goodness of fit indices are used to assess the fit of the model. In the present study, χ2/df, GFI, CFI, and RMSEA were utilized. To have a good fit model, χ2/df should be less than 3, GFI and CFI should be above .90, and RMSEA should be less than .08 (Kunnan, 1998).

A model based on the three factors of politeness scale was tested. There were 22 items in the initial model-HPS= 8 items, DPS=4 items, and SPS= 10 items. The initial model with all the items did not fit the data well, so some modifications were applied. All the items which had non-significant factor loadings were removed (h1, h4, s1, s2, s4, & s7). The final result can be seen in Figure 1.

Figure 1.The results of CFA for the CLP scale

The final model showed a moderate fit to the data (chi-square/df = 2.82, GFI = .91, CFI = .84, RMSEA = .07). Then the reliability was checked with Cronbach's alpha. The step-by step modifications can be seen in Table 1.

Table 1

 Step-by-Step Modifications of the CLP Scale

Results showed the following alphas for the subscales of politeness:  HPS= .83, SPS=.88, DPS= .81.

In order to assess Iranians’ conceptions of politeness, mean of the three politeness conceptions was taken into account. Mean of the three strategies conceptions can be seen in Table 2.

Table 2

 Descriptive Statistics for the CLP Scale

As Table 2 indicates, DPS has the highest mean (M= 3.56) among the three politeness conceptions and SPS (M= 2.94) and HPS (M= 2.87) have almost the same mean. Therefore, Iranians use this politeness system more than the other two types.

5.2. Validating the DCP Scale

In order to assess the validity of the research scale, CFA was used. A model based on the two factors of the scale was tested.

The initial model comprised of 19 items-explicit=9 items and implicit=10 items. The initial model with all the items did not fit the data well, so some modifications were applied. All the items which had non-significant factor loadings were removed. In accordance, items1, 14, and 7 were omitted from explicit subscale and items 4,9,16, and 17 were omitted from the implicit subscale. The final result can be seen in Figure2.

Figure 2. The results of CFA for the DCP scale

The final model showed good fit to the data (chi-square/df= 2.21, GFI= .95, CFI= .94, RMSEA= .07). The step-by step modifications can be seen in Table 3.

Then the reliability was checked with Cronbach's alpha. Results showed the following alphas for the subscales of politeness:  implicit= .89, explicit= .86.

Table 3

Step-by-step Modificationsthe DCP Scale

In order to assess the Iranians’ data collection preferences, mean of the two research subscales (implicit and explicit) was taken into account. Mean of the two research subscales can be seen in Table 4.

Table 4

 Descriptive Statistics for DCP Subscales

As Table 4 shows, implicit items have a higher mean (M= 26.74) compared to explicit items (M=19.40). Therefore, while taking part in research, Iranians prefer to answer the implicit items more than the explicit ones.

5.3 Correlation

In order to assess the relationship between the implicit and the explicit subscales of DCP with CLP subscales, Pearson product-moment correlation formula was used. The results can be seen in Table5.

Table 5

 The Relationship between the DCP and the CLP Scales

** The mean difference is significant at the 0.01 level.

* The mean difference is significant at the 0.05 level.

As Table5indicates, there is no significant relation between the explicit subscale of DCP with all sub-scales of CLP: HPS (r= .08, p> 05), SPS (r = .05, p>.05), and DPS (r= -.03, p>.05). Then, the relationship between the implicit subscale of DCP and CLP subscales was examined. As Table 5 shows, there is a significant relation between the implicit subscale of DCP and all sub-scales of CLP: HPS (r= .20, p<.01), SPS (r= .23, p<.01), and DPS (r= .17, p<.05). Among the subscales of politeness, SPS (r= .23, p<.01) had the highest correlation with the implicit subscale of DCP.

Then, the relation between each single item of explicit and implicit DCP subscales was examined with CLP subscales. First, explicit items were examined.

Table 6

 The Relationship between Explicit Items of the DCP and the CLP Scales

As Table 6 shows, there is no significant relation between any explicit items of the DCP scale with CLP scale. Subsequently, the relations between implicit items of DCP scale and the CLP scale were examined. 

Table 7

 The Relationship between Implicit Items of the DCP and the CLP Scales

As Table 7 indicates, there is a significant relationship between item 19-indirect observation- (r= .14, p<.05) with HPS. Besides, there is a significant relationship between item 12-metaphorical items- (r = .22, p<.01) with SPS. Finally, item 5- unstructured interview- (r = .14, p<.05) was found to have a significant relationship with DPS. No other significant correlations were found.

6. Discussion

Substantial amount of research has been performed on discovering cultural patterns (Hofstede, 2001; Klein, 2003) and on the way cultural factors affect cognitive processes (Nisbett, 2003). In the same vein, it seems that data collection is under the influence of cultural factors such as politeness (Pishghadam, 2013), and there is mounting evidence that, although there may be some common underlying features, politeness is perceived differently across cultures (Almursy & Wilson, 2001; Ide, 1989; Janney & Arndt, 1993; Lee-Wong, 2002; Matsumoto, 1989; Nwoye, 1989). Nevertheless,no validated scales have been designed for data collection preferences as well as conceptions of linguistic politeness in Iranian context. Therefore, to examine the relationship between data collection and linguistic politeness, this study attempted to devise and validate a research scale operationalizing Iranians’ data collection preferences. In parallel, the present study tried to devise and validate an emic scale operationalizing Iranians’ conceptions of linguistic politeness. Devising such scales through which CLP as well as DCP are derived would be of overriding importance as they are specifically designed and localized for the Iranian context. Finally, the relationship between the CLP and the DCP scales are examined.

In order to assess Iranians’ data collection preferences, mean of implicit as well as explicit DCP subscales were taken into account. Results showed that the implicit items a higher mean compared to the explicit items.  So, while taking part in research, Iranians prefer to answer the implicit items more than the explicit ones. The justification of this finding can be that, adhering to high-context communication is mostly seen among people in collectivistic cultures, while using low-context communication is common among people in individualistic cultures (Gudykunst & Ting-Toomey, 1988). Low-context communication focuses on direct verbal messages, while high-context communication emphasizes indirect verbal messages and nonverbal messages (Hall, 1976). Consequently, collectivism with the feature of low-context communication has made Iranians prefer to be asked indirect and implicit questions rather than explicit and direct ones when they take part in research.

In the same vein, in order to assess Iranians’ conceptions of linguistic politeness, the means of the three CLP subscales were taken into account. Deference Politeness System (DPS) has the highest mean among the three politeness subscales-DPS, HPS,& SPS-,thus Iranians tend to use this politeness system more than the other two types. Since individualism-collectivism is tied with cultural differences in managing face (Imahori, & Cupach, 1994; Ting-Toomey & Kurogi, 1998), and cultural individualism-collectivism directly influences cultural rules (Triandis, 1995), the ways members of a culture perceive politeness rules can be varied.

In the Iranian context, there is less emphasis on individuality compared to Westerners, and collectivity or group-identity rather than individuality or self-identity is prioritized (see Kinnison,2000). Group consciousness and interdependence set the main aspects of social relationships in the Iranian culture (see Koutlaki, 2002). As a result, horizontal collectivism stressing on interdependence, sociability, and common goals (Singelis, Tdandis, Bhawuk, & Gelfand, 1995) seems to be more favored in Iran than vertical collectivism in which hierarchical relationship is highlighted (Chen, Meindl, & Hunt, 1997).

Furthermore, Iranians struggle to be polite and show deference towards others according to cultural and social schema of face "âberu" and set it as their first priority in each and every aspect of their life (Sharifian, 2007).  Aberu, or honour, as a powerful social force (O’Shea, 2000), entailing not only family possessions, appearance, etc., but also concerning about one's behavior, actions, and social identity, has compelled Iranians to maintain their face everywhere.

Finally, with respect to the last objective of the study, the relations between the implicit and the explicit DCP subscales with CLP subscales (HPS, DPS, &SPS) were checked using Pearson product-moment correlation formula. The results suggested no significant relation between the explicit subscale of DCP with all sub-scales of CLP, while there was a significant relation between the implicit subscale of DCP and all sub-scales of CLP. Therefore, it seems that Iranians’ data collection preferences are under the influence of their conceptions of linguistic politeness. That is to say that, in the present study, the more polite people’s conceptions get, their tendency to answer the implicit items in research increases. Therefore, among the cultural factors affecting cognitive processes (Nisbett, 2003), politeness conceptions deserve meticulous attention, as in a collectivistic and high-context society such as Iran, deductive focusing on relationship, synthetic reasoning, introspection, indirect, and non-linear thinking is highlighted (Hall, 1976).

Furthermore, examining the relation between each single item of the explicit and the implicit subscales of DCP with CLP subscales revealed that, there is no significant relation between any explicit items of DCP with CLP subscales, while there is a significant relationship between indirect observation with hierarchical politeness system. The assumption underlying this method of data collection is that the behavior under study is purposeful and entails deeper values and beliefs. Observation may vary from a highly structured, specified description of behavior organized by checklists (direct observation) to a more holistic notation of events and behavior (indirect observation) (Marshall & Rossman, 2010). Since indirect observation makes use of no predetermined observational checklist, the researcher has more elbow room to discover patterns underlying the studied behavior as well as relationships and complex interactions among variables (Marshall & Rossman, 2010).  Therefore, in a collectivistic society such as Iran, people are more willing to take part in indirect observation as its features are in harmony with their culture focusing on relationship, synthetic reasoning, introspection, indirect, and non-linear thinking (Hall, 1976).

In addition, there is a significant relationship between metaphorical items -the highest correlation- with SPS.A metaphor is defined as a figure of speech associating something with an unrelated thing in order to emphasize the similarities between them. Since Iran is a high context culture focusing on indirect verbal messages (Hall, 1976) on the one hand, and there is a relationship between the level of indirectness and politeness (e.g. Brown & Levinson, 1987; Leech, 1983) on the other, indirect data collection methods such as metaphorical analyses are suitable to be used in the Iranian context (Pishghadam, 2013).

Finally, unstructured interviewhad a significant relationship with DPS. Interviewing is described as a conversation which has a purpose (Kahn & Cannell, 1957). It may vary in terms of a priori structure (structured)or the interviewee’s responses to questions (unstructured or open-ended). In unstructured interview, the researcher seeks to uncover the participant’s views while respecting the participant’s framing and structuring of the responses, and conveying the attitude that his/her views are valuable and beneficial (Marshall & Rossman, 2010). The underlying assumption in unstructured interview is that the participant’s point of view about the subject of interest should unfold as the participant views it- the emic perspective- rather than etic perspective formed by what the researcher views about the phenomenon of interest (Marshall & Rossman, 2010). The participants of the present study preferred to answer unstructured interview questions as in a collectivistic high- context society like Iran, again, indirect way of thinking, reasoning, and introspection in order to observe politeness are highlighted.

The main implication of the findings of the present study is that western type of data collection should be used with great caution in Iran (Pishghadam, 2013), since these instruments have been validated in the west according to their individualistic, low-context, and inductive culture which has emphasis on categorization, analytic reasoning, self-disclosure, retrospection, direct, and linear thinking (Hall, 1976). On the contrary, ina collectivistic and high-context culture such as Iran, indirect verbal message, non-linear thinking, relationship, and interdependence are highlighted to observe politeness rules. Accordingly, using the instruments which are validated with etic approach criteria don't seem to be a proper and valid means of data collection in Iran, as the findings achieved through them are assumed to be affected by the conceptions of politeness among Iranians. Thus, based on the aforementioned findings, it seems reasonable for researchers to look for instruments localized for their own culture or at least try to implement etic standardized instruments with meticulous care by modifying and adapting their content to the features of their national culture.

The main attainment of the present study is to devise and construct two emic scales for Iranians’ conceptions of linguistic politeness as well as Iranians’ data collection preferences. Besides, the relations between these two scales have been studied to shed light on the use of emic data collection methods in Iran. However, there were some limitations in the present study. Researchers used Scollon and Scollon’s (1995) theory of politeness systems for devising the CLP scale. Other theories or methods could be utilized for devising the CLP scale such as in-depth interview, observation, field-notes, etc. Moreover, in this study Amos 20 was utilized to check the construct validity of the scales using CFA method. Future research can use other methods such as Item Response Theory (IRT) for validation process.