Introduction

“Where ever you go, there you are”—this simple and seemingly trivial statement of Kabat-Zinn captures the essentials of mindfulness in seven words (Kabat-Zinn 2005). Mindfulness can be understood as the ability to dispassionately observe the experience of the present moment with non-judgmental openness. Probably the oldest written reference to mindfulness is the notion of sati in the Buddhist Pali canon (Schmidt 2011). Buddhist scholars explain that term as “mere observing”—just observing an object without interfering (Thera 1993); see also Analayo’s excellent treatise on sati for a more detailed analysis (2003). Bishop et al. (2004) proposed a two-factor model of mindfulness consisting of an attentive factor focusing on present moment experience and an emotion regulation factor characterized by openness, curiosity, and acceptance. On the contrary, Brown and Ryan (2003) conceptualize mindfulness as a unidimensional construct where attentive and emotive factors cannot be disentangled. As will be laid out in this contribution, this difference reflects one of the major points of debate with regard to understanding, conceptualizing, and thus also operationalizing mindfulness.

There is no doubt about the fact that mindfulness implies a sense of being in the present, vivid awareness of sensory and mental experience, and calls for a voluntary disruption of automatic mental processes of cognitive evaluation such as “good,” “bad,” “wanted,” or “unwanted.” Additionally, many spiritual traditions as well as contemplative and meditative schools from various contexts emphasize the need for and the use of being mindful. Nevertheless, mindfulness is usually closely related to Eastern traditions and predominantly Buddhist philosophy (Schmidt 2011). However, concepts similar to mindfulness can also be found in several Western philosophical systems, for example, in the work of Husserl or Buber (Buber 2004; Kockelmans and Husserl 1994). According to Husserl’s phenomenology, in order to achieve profound knowledge of an object, this object needs to be scrutinized objectively, the respective observer thereby attempting to not rely on presuppositional knowledge. In its essence, Husserl’s conceptualization of phenomenology aims at establishing what can be termed a scientific analysis of consciousness and related subjective experiences. In contrast to other scientific enterprises, however, it uses subjective methods in an attempt to describe an experience and the structure constituting the experience (Sepp and Embree 2010). Buber’s philosophy, foremost his “I–Thou” concept, embraces interpersonal domains by describing phenomenal percepts (i.e., qualia) as being associated with an intense encounter and exchange of two human beings. To be able to enter such an “I–Though” encounter, a mindful interaction of two subjects is necessary. This encounter is characterized by a mutual stance of not thriving for anything particular in the dialogue, but rather for accepting the other person and the situation as it is. Such situations are, according to Buber, not only rich and fruitful encounters but actually the only possibility to experience the existence of another person as a “real” person (Friedman 2002). Additionally, psychologists who understand mindfulness predominantly from an information processing point of view, have also contributed substantially to the understanding of mindfulness. For example, Langer’s (1989) concept of mindfulness focuses on drawing novel distinctions on the information derived from perception. Deriving new categories is what keeps one situated in the here and now. Further aspects of Langer’s concept of mindfulness are openness to new information, flexibility to take over different cognitive perspectives, and a “step by step” attitude, so that one concentrates on the task at hand. Langer’s concept overlaps with Eastern concepts of mindfulness, e.g., as explained by Kabat-Zinn (1990). However, Langer’s focus on novelty which resembles the construct of creativity is at a remarkable distance to the Eastern concepts of mindfulness.

Most authors would—despite different conceptualizations—probably opine that mindfulness is an inherently human quality. While there may be substantial inter- and intra-individual differences in the natural degree of mindfulness (Brown and Ryan 2003), research has shown that it can be developed with training (Davidson et al. 2003; Shapiro et al. 2008).

In recent years, mindfulness as a psychological function has not only received considerable research interest but has also been frequently applied in clinical contexts for ameliorating a variety of conditions such as anxiety and personality disorders, substance abuse, stress, as well as chronic pain, to name just a few. To date, several reviews aggregate the findings and, over and above, substantiate the clinical effectiveness (e.g., Burke 2010; Chiesa et al. 2011; Fjorback et al. 2011; Hofmann et al. 2010; Mars and Abbey 2010; Walach et al. 2012). For example, Bowen et al. (2006) showed that in a sample of 173 inmates of a US prison (57 in the mindfulness group, 116 in the treatment-as-usual control group), individuals of the mindfulness showed less substance abuse in alcohol, marijuana, and crack cocaine after the training. Additionally, mindfulness-group individuals showed less psychiatric symptoms and increases in positive psychosocial outcomes. The mindfulness-group was trained in Vipassana meditation, a Buddhist mindfulness meditation technique. Results from brain imaging studies have found a positive relationship between mindfulness and neural functions associated with well-being (Farb et al. 2007; Tang et al. 2007, 2009), while effects on immune parameters have also been reported (Carlson et al. 2004; Davidson et al. 2003; Witek-Janusek et al. 2008).

In summary, there are a substantial number of studies affirming the positive health effects of mindfulness, although it has to be noted that the magnitude of these effects and the methodological quality varies considerably between studies. Furthermore, a number of apparently influential treatment approaches have been presented, which have in turn certainly exhibited influence upon the development of some of the measurement instruments developed for assessing mindfulness. Among these interventions, the Mindfulness-Based Stress Reduction Program, originally developed by Kabat-Zinn (1990), is probably the most well-known method (Irving et al. 2009). Additionally, Mindfulness-Based Cognitive Therapy, based on the work of Segal et al. (2002a, b), Dialectical Behavior Therapy, based on Linehan (1993), and the Acceptance and Commitment Therapy, based on Hayes et al. (1999), have frequently been used in clinical practice (Didonna 2008). However, despite the high interest in the construct of mindfulness, doubts regarding the assessment of this construct remain (Grossman 2008, 2011). For example, mirroring an important point of the conceptual debate associated with mindfulness, some questionnaire instruments operationalize mindfulness as a unidimensional construct, whereas others argue for a two- or multifactorial structure (see section on “Conceptual Frameworks” for details).

Hence, given the popularity of mindfulness and the concerns regarding the adequate measurement of it, it seems clear that current approaches to measuring mindfulness should be examined systematically in a bid to further develop and refine the measurement of mindfulness. Until now, mindfulness has predominantly been assessed by one or more existing psychometric scales published in peer-reviewed journals; see Didonna (2008) and Baer (2011) for an overview of some of the scales. Some alternatives to psychometric scales such as interviews have been proposed in order to compensate the shortcomings of questionnaires (Grossman 2008). However, a concise summary of the existing mindfulness measures—both psychometric and otherwise—that summarizes the approaches and contrasts the particular strengths and weaknesses is missing.

Assessing Self-Assessment of Mindfulness on the Basis of Psychometric Scales

Utilizing psychometric scales is until now the standard approach for measuring self-attributed mindfulness state and trait levels. Consequently, a substantial and growing number of scales have been published (see Table 1). The advantages of the psychometric approach are straightforward: convenient and quick application, well-known methodology, and empirical support (Baer et al. 2004; Bohlmeijer et al. 2010; Chiesa et al. 2011; Mars and Abbey 2010). At the same time, there are shortcomings which may considerably limit the validity of this method.

Table 1 Background information and calibrated populations
Full size table

Grossman (2008) has brought forward a twofold criticism. First, he points out a lack of common ground regarding the definition of mindfulness. This is, according to Grossman, largely due to the fact that some authors are not suitably familiar with the theoretical basis of the concept of mindfulness as, for example, laid out in Buddhist psychology. Grossman has additionally argued that there is not a single coherent concept of Buddhism but that the interpretation of mindfulness rather varies between different Buddhist traditions (e.g., Theravada and Tibetan). We agree with Grossman and opine that conceptual pluralism is probably not only acceptable but, from a pragmatic point of view, actually inevitable. This is because mindfulness, although it was introduced and made predominant by Buddhist teachings, is likely to be a universal psychological function, practice, or phenomenon that is dependent on culture and context. We therefore believe that theoretical approaches stemming from different cultures and contexts may actually not contradict each other but actually add to the exploratory fundament that is necessary to grasp all facets of mindfulness (at least if mindfulness is not only considered to be an intrapersonal but also interpersonal construct).

It should additionally called to mind that mindfulness is a construct that may seem alien to common causal deterministic thinking, as mindfulness inherently involves a dialectical structure (Sauer et al. 2011a). For example, mindfulness combines active (actively attending to what is happening) and, at the same time, also passive (not reacting in the face of unwanted stimuli) components in a complementary manner. Furthermore, it has been argued that mindfulness does not entail striving for symptom alleviation but is rather a world view and a way of being; symptom alleviation may occur as a by product or candidly as positive collateral damage (Sauer et al. 2011a).

Grossman’s second criticism hinges on the fact that by their very nature, self-report questionnaires rely on introspective self-perception. Gauging one’s own mindfulness level may be biased due to, for instance, an idiosyncratic and rather naive understanding of mindfulness. As Grossman argues, it may well be the case that both a novice and highly proficient meditator rate their ability to be mindful as medium, given that the novice is not familiar with the concept and the novice has become humble during his practice given that she becomes aware of her mindlessness. This is not merely a theoretical argument; as in a study by Leigh et al. (2005), the authors found that binge drinking students rated themselves as significantly more mindful than their abstinent peers. Hardly anyone would argue that from a theoretical point of view binge drinking may serve as an indicator for mindfulness.

Nevertheless, this is not a phenomenon unique to mindfulness but is, for example, well-known in quality of life research as “shifting baseline” or “response shift” (Güthlin 2004; Oort et al. 2009), which occur due to a change in internal reference standards. Thus, some practice in mindfulness may help a practitioner realize how less mindful he or she actually is and hence the self-assessment of mindfulness levels of such a more advanced practitioner might, paradoxically, be lower than that of a novice. Additionally, other effects well-known in social psychology may occur such as consistency effects (respondents may try to answer a second item in a consistent way to a first one even if their opinion is in reality different) or social desirability effects (respondents may try to present themselves in a more positive, i.e., mindful, way even if this self-presentation does not reflect their true attitude). In a recent update of his criticism, Grossman (2011) adds that evidence in favor of the scales is still scarce, particularly regarding the translation of psychometric data into behavioral data. Whereas a substantial number of studies indicate that psychometric approaches may be seen as legitimate venues for assessing mindfulness, we agree with Grossman that studies linking psychometric and behavioral data are for the time being thinly distributed—but many other psychological research areas are also beset by this problem (Brown et al. 2011a; but cf. Sauer et al. 2011c).

Each of these points of criticism raised by Grossmann has nevertheless been taken seriously as these shortcomings may severely impede the validity and usefulness of self-report instruments. We reiterate that these effects are not unique to mindfulness research but may actually limit the usefulness of self-report questionnaire instruments in general and in many other research domains. Additionally, from a pragmatic point of view, even though self-report measurements are, by their very nature, imperfect, the vast majority of published studies speak in favor of theory conform measurement effects of mindfulness, as documented by several reviews and meta analyses (Bohlmeijer et al. 2010; Chiesa et al. 2011; Mars and Abbey 2010). Evidence regarding social cognition effects such as social desirability on the measurement of mindfulness is to our knowledge still scarce. In the few studies addressing this issue, no substantial association of mindfulness with social desirability has been detected as yet (Brown and Ryan 2003).

Our aim is to (1) describe the existing approaches of measuring mindfulness, (2) evaluate the strengths and weaknesses of these approaches, and (3) point out research avenues to advance the assessment of mindfulness including recommendations regarding the suitability of certain instruments in given research contexts.

Method

To identify relevant research articles, a literature search was conducted using MEDLINE (PubMed), PsycInfo (EBSCO), and ScienceDirect databases. Hand searches of references in the retrieved literature were additionally performed. Our search for studies tapping into the measurement of mindfulness included articles and empirical work published until July 17, 2011. The search strategy considered only studies published in English. Primary search terms were “mindfulness,” “measurement,” “assessment,” “questionnaire.” We performed an additional search on predictive validity as it is a strong criterion of the quality of a psychometric scale. Secondary search terms included mindfulness, the name of the respective scale either spelled out in full or abbreviated, and the name of the first author. To get an estimate of the prosperity of each scale, we counted the number of citations for the initial validation article (see Table 2) using Google Scholar as an index measure (retrieved on July 16, 2011). Scales focusing on particular subsets of behavior such as eating were excluded (Framson et al. 2009), and scales for particular, non-generic populations such as adolescents were also not considered (Greco et al. 2011).

Table 2 Literature database search for number of citations for mindfulness scales
Full size table

Results

Unpublished questionnaires are sporadically mentioned in the literature (Bodner and Langer 2001; Chadwick et al. 2005; Kee and John Wang 2008). However, we confined this review to published material in PubMed and PsycInfo. As a result, a number of adaptations of existing scales will not be examined (Brown et al. 2011b; Höfling et al. 2011a, b; Jimenez et al. 2010). Furthermore, we excluded questionnaires focusing on non-generic but rather specific populations, such as adolescents (Brown et al. 2011b), and special behavior set such as eating (Framson et al. 2009). Applying our research strategy with the exclusion criteria just described, we were able to identify eleven instruments:

  1. 1.

    Cognitive and Affective Mindfulness Scale Revised (CAMS-R) (Feldman et al. 2007): According to the authors, the questionnaire was “designed to capture a broad conceptualization of mindfulness” (p. 178; Feldman et al. 2007), based on the work of Kabat-Zinn (1990). The authors provide initial support for their scale consisting of four factors (attention, present focus, awareness, acceptance) and a second-order mindfulness factor. A short version of the scale is also presented in the development article (Feldman et al. 2007). Sample items include: “It is easy for me to concentrate on what I am doing” and “I am preoccupied by the future” (reversed).

  2. 2.

    Developmental Mindfulness Survey (DMS) (Salloway and Fischer 2007): This is the only scale built on an explicit effort to use item response theory approaches for the development of a mindfulness scale. Mindfulness is conceptualized as a unidimensional construct based on a concept of mindfulness consisting of open awareness of the present and a non-judgmental stance (Solloway 2001). An interesting feature of the methodological approach used by the authors of this scale is that they defined an a priori difficulty ranking of the items, i.e., easy, intermediate, and difficult items. Evaluation of mindfulness training thereby allows for testing the presupposed item difficulty ranking. Sample items include: “I am learning that paying attention to what is happening right now is harder than I first thought it would be” (low difficulty), “As I practice paying attention to what is happening right now, I notice that my mind wanders frequently” (intermediate difficulty), and “When I practice paying attention to what is happening right now, I observe my thoughts without being caught up in them” (high difficulty).

  3. 3.

    Effects of Meditation Scale (EOM) (Reavley and Pallant 2009): This scale is different to the rest of the scales presented here as it encompasses mindfulness as a broader construct in the context of meditation and meditative experiences. Therefore, in addition to mindful awareness, specific effects of meditative practice are additionally assessed. The instrument consists of two subscales: experiences during meditation (EOM-DM) and effects of meditation in everyday live (EOM-EL). Sample items include: “I experience fluctuating emotions” (EOM-DM), “I make more effort to avoid causing suffering to others” (EOL-EL).

  4. 4.

    Five Factors Mindfulness Questionnaire (FFMQ) (Baer et al. 2006): The FFMQ is an attempt to unify the most widely used mindfulness questionnaire in one tool. The resulting instrument is similar to the Kentucky Inventory of Mindfulness Skills (KIMS) (Baer 2004) but includes one more factor (in total: nonreactivity to inner experience, observing/noticing/attending to sensations/perceptions/thoughts/feelings, acting with awareness/automatic pilot/concentration/nondistraction, describing/labeling with words, nonjudging of experience). Mindfulness is conceptualized as a general second-order construct that related to the five specific first-order factors. Items are drawn mainly from the KIMS.

  5. 5.

    Freiburg Mindfulness Inventory (FMI) (Buchheld et al. 2001; Kohls et al. 2009; Walach et al. 2006): This scale explicitly builds on the premises of Buddhist psychology. Recent research favors a two-dimensional factorial structure with an attention to the present moment aspect (factor presence) and a factor of non-judgmental attitude (factor acceptance). Besides the initial 30-item version, a 14-item version, and a 7-item version have been proposed (Jimenez et al. 2010). Sample items include: “I am open to the present moment”(presence factor), “I am able to smile when I notice how I sometimes make life difficult” (acceptance factor).

  6. 6.

    Kentucky Inventory of Mindfulness Skills (Baer 2004): The KIMS tries to incorporate mindfulness skills as taught by well-known mindfulness-based interventions such as mindfulness-based stress reduction (MBSR) (Kabat-Zinn 1990), mindfulness-based cognitive therapy (Segal et al. 2002a, b), Acceptance and Commitment Therapy (Hayes et al. 1999) and Dialectical Behavioral Therapy (Linehan 1993). The scale consists of four factors (observe, describe, act with awareness, nonreactive stance). A short form of the scale has recently been published (Höfling et al. 2011b). Sample items include: “I am able to pay close attention to one thing for a long period of time” (factor act with awareness), “I watch my feelings without getting lost in them” (factor nonreactive stance).

  7. 7.

    Langer Mindfulness/Mindlessness Scale (Haigh et al. 2011): Interestingly, the authors of this scale conceptualize mindfulness on the basis of information processing and creativity theory. In contrast to other scales, this instrument is not explicitly based on Buddhist ideas but rather reflects a Western approach to mindfulness that originates in cognitive psychology. Sample items include: “I like to investigate things,” “I am very curious.”

  8. 8.

    Mindfulness Attention Awareness Scale (MAAS) (Brown and Ryan 2003; Höfling et al. 2011a): The MAAS is probably the most widely used scale to date, despite recent criticism (Grossman 2011; Grossman and Van Dam 2011). Whereas the majority of the other scales define mindfulness as embracing both an emotional and an attentional aspect, the MAAS conceives mindfulness to consist only of an attentional aspect. Additionally, this scale does not measure mindfulness but rather “mindlessness”—assuming that mindlessness can be seen as the inverted concept of mindfulness (i.e., the two constructs can be converted by simply recoding the items). An alternative and shorter version has recently been proposed (Höfling et al. 2011a). Sample items include: “I break or spill things because of carelessness, not paying attention, or thinking of something else” (reversed), “I drive places on “automatic pilot” and then wonder why I went there” (reversed).

  9. 9.

    Philadelphia Mindfulness Scale (PHLMS) (Cardaciotto et al. 2008b): The PHLMS has a clear-cut factorial structure with two factors (awareness, acceptance) building on Bishop’s et al. (2004) concept. Mindfulness is not seen as a second-order factor but rather as consisting of two independent but interrelated factors. Sample items include: “I notice changes inside my body, like my heart beating faster or my muscles getting tense,” and “When I have a bad memory, I try to distract myself to make it go away” (reversed).

  10. 10.

    Southampton Mindfulness Scale (SMQ) (Chadwick et al. 2008): The SMQ supports a unidimensional structure of mindfulness but includes items pointing to present moment awareness and items pointing to accepting attitude towards all experience. The scale focuses on mindfulness towards unpleasant thoughts and pictures. Sample items include: “Usually when I experience distressing thoughts and images… I am able to notice them without reacting,” and “…I am able to accept the experience.”

  11. 11.

    The Toronto Mindfulness Scale (TMS) (Lau et al. 2006): The TMS measures mindfulness as a state, and not as a trait, in contrast to all other scales described here. The conceptual base is built on Kabat-Zinn’s (1990) work. This measurement approach opens up a new research perspective as it was designed to measure state mindfulness in a given situation but has until now not been frequently utilized. Sample items include: “I noticed subtle changes in my mood,” “I was more concerned with being open to my experiences than controlling or changing them.”

The psychometric properties of all questionnaires are depicted in Tables 1 and 3. Each scale is presented in the following format: description of the theoretical background, author of development article, conceptualization of mindfulness as trait or state, types of samples used for development, scale format details, and versions and adaptations.

Table 3 Scale format details
Full size table

Discussion

Conceptual Framework of the Scales

Current conceptual challenges of the scales may be summarized in three points. Some scales have explicitly been constructed on the basis of a clear fundament. For example, the FMI explicitly draws explicitly on Buddhist psychology, while the MMS was developed on the basis of information processing theory; some other scales such as the KIMS are based on a more eclectic and pragmatic approach. We opine that an explicit theoretical grounding in a conceptual framework such as Buddhist psychology or information processing theory is a desirable property of any model or measurement instrument, as this allows gauging the degree of alignment of the scale with a specific conceptual background framework (i.e., conceptual validity).

Second, similar to the critique brought forward against measurement instruments of spirituality by Koenig (2008), there is a danger that mindfulness questionnaire items are designed in a way that they are contaminated with variables relevant for health and well-being while solely asserting to capture mindfulness. For example, items such as “I experience moments of inner peace” may possibly be seen as a generic indicator for mental health but not necessarily be related with the ability to be mindful. If mindfulness as a mental process, state or practice is intermingled with health and well-being parameters, empirically observed associations between mindfulness and health parameters have to be considered to be meaningless and tautological. We therefore believe that sticking to the behavior, attitudes, and cognitions associated with mindfulness derived from one explicit theoretical background is a way to develop specific items without falling prey to broader and less specific levels of descriptions that may actually blur the construct.

Third, most authors seem to agree—explicitly or implicitly—that mindfulness consists of two distinct factors. The first factor can be conceived as presence, an attentional focus to momentary experience, whereas the second factor can be conceived as acceptance, a genuine non-judgmental, open, and accepting attitude towards what is happening in the present moment (Bishop et al. 2004). Some scales do explicitly draw on these two factors (e.g., PHLMS, TMS), whereas other scales draw on other subfactors, which nevertheless may be allocated to the two latent factors presence and acceptance denominating both a cognitive and an emotional component of mindfulness. Recent findings concerning the FMI provide, for example, not only psychometric but actually experimental evidence for the validity of the “two-factor” concept of mindfulness (Kohls et al. 2009; Sauer et al. 2011b).

Scales conceptualizing mindfulness as a unidimensional construct are, for example, the DMS and the MAAS. The authors of the MAAS hold that the acceptance factor is from a psychometric point of view redundant as the acceptance facet failed to account for additional variance in factor analysis of their data set (Brown and Ryan 2004). A possible explanation for this finding may be that the MAAS has been built on a different, but not necessarily incompatible, unidimensional conception and definition of mindfulness that focuses on the presence factor. It is well conceivable that mindfulness refers to nothing else than simply being aware of the present moment, as advocated by Brown and Ryan. According to this “purist” view, mindfulness may be considered to be nothing other than perceiving the present moment as it is, while judgmental and evaluative thinking has been suspended as a consequence of present moment awareness. Correspondingly, as no meta-cognitive processes such as evaluative reasoning or reacting are involved, one may say that only present moment awareness resides (assuming that acceptance is a prerequisite for being present).

On the other hand, supporting the two-dimensional conceptions with present moment awareness and acceptance as the two central but interrelated aspects of mindfulness, it should be noted that in normal states of mind, emotional behavioral reactions automatically occur. Hence, intentional effort is needed to stop these automatic and to a large degree also implicit processes. Seen from that angle, it seems reasonable to state that one needs to systematically train the ability to actively inhibit cognitive evaluative processes. This idea is in line with those authors who endorse acceptance as the second key component of mindfulness. With increasing mindfulness, suppressing or at least voluntarily decreasing cognitive evaluative processes is thought to become less difficult, as automatic and implicit processes of evaluating are partially decoupled from their respective percepts. Thus, with increasing proficiency in mindfulness, one may indeed feel that being mindful is primarily associated with the ability to pay attention and to focus awareness in an effortless way, and that restraining secondary cognitive processes as such does not require extra effort.

On the contrary, novices frequently report to feel the need to voluntarily constrain judgmental processes. Thus, whether or not the training and conscious execution and maintenance of acceptance can be seen as an additional facet of mindfulness or as a prerequisite for being in a state of pure presence might be dependent on the proficiency of the practitioner. Put bluntly, the question boils down to the well-known hen or egg problem: Is acceptance a prerequisite for presence or does the ability to be present allow to develop acceptance? Initial data support the latter hypothesis. For instance, Kohls et al. (2009) suggest that presence may be interpreted as a causal mechanism that builds acceptance, and presence is therefore expected to be developed earlier than acceptance. Brought together, theoretical arguments for both the unidimensional as well as the two-dimensional conception of mindfulness can be posited.

If mindfulness could be shown to be associated with effortful and voluntary suppression of evaluative metacognitive processes—especially at an early stage in mindfulness training—it would further substantiate the claim that two processes are at work, namely presence and acceptance. In contrast, if apart from attentional processes no other specific cognitive or emotional self-regulation processes can empirically be observed, the claim of Brown and Ryan indicating that mindfulness is solely an attentional phenomenon would be supported.

Empirical Support of the Scales

The empirically derived psychometric properties of the scales are comparable as most scales have been validated by a number of studies utilizing different types of study designs and methodologies. Yet it should be noted that the number of studies employing a certain scale differ greatly (see Table 2). Each study utilizing a certain mindfulness scale and confirming initial hypotheses about associations of mindfulness with other parameters can, in principle, be seen as adding weight to the claim that this scale is a valid instrument. Thus, although this cannot be considered to be the sole criterion, given the self-stimulating dynamics of positive feedback loops, the number of studies having successfully employed a certain scale can—cum grano salis—be interpreted as a sign of empirical support. Following this logic, we counted the citations for the article describing the initial development and validation of a respective scale. Results suggest that the KIMS and the MAAS are, for the time being, the most widely employed scale for measuring mindfulness, followed by the FFMQ (see Table 2). Although the figures are not standardized by publications per year, results still allow gauging the general popularity of the scales despite the uncorrected age bias (i.e., formerly developed scales have an advantage over recently developed instruments).

Most scales have been validated by means of confirmatory factor analysis and showed good psychometric properties such as internal consistency. Secondly, all scales with the exception of the CAMS-R and the MMS have been validated in different samples, while the CAMS-R and the MMS have until now only been validated in student samples.

Predictive validity, i.e., predicting future trajectories of a given variable, is a useful indicator for assessing the general validity of a scale. However, predictive validity has been tested, to a much lesser degree, compared to cross-sectional correlation analyses. Comparing the number of studies that explicitly tested the predictive validity of mindfulness scales, we found that for the time being, the MAAS receives the strongest support. Its predictive validity was supported by nine studies (Barnes et al. 2007; Gilbert and Christopher 2009; Levesque and Brown 2007; Mitmansgruber et al. 2009; Nyklícek and Kuijpers 2008; Shapiro et al. 2008; Van Dam et al. 2010; Vidrine et al. 2009; Weinstein et al. 2009). For the KIMS, we found three studies corroborating its predictive validity (Nyklícek and Kuijpers 2008; Ong et al. 2009; Vujanovic et al. 2010). For the PHLMS, we found one study (Cardaciotto et al. 2008a); for the FFMQ, we found two studies (Carmody and Baer 2009; Fink et al. 2009), for the MMS, we found one study (Kee and John Wang 2008), as well as for the FMI, for which we found one study corroborating its predictive validity (Huffziger and Kuehner 2009). For the rest of the mindfulness scales, we could not retrieve evidence for their predictive validity.

A major drawback of most scales is that the underlying measurement models of the majority of the scales have not yet been tested, with the exceptions of the DMS (Salloway and Fischer 2007), FMI (Sauer et al. 2011b), and the MAAS (Höfling et al. 2011a). Such analyses can be performed using item response theory models such as the Rasch model (Rasch 1960). Item response theory analysis is a psychometric approach complementing traditional approaches that is increasingly acknowledged as a complementary approach to classical test theory (Fischer and Molenaar 1995). By way of item response theory, the central assumption can be tested whether the convenient practice of collapsing item values into a sum score is justified. In contrast, in the case of classical test theory, item difficulties and person parameters are mutually confounded. As a result, items of high difficulty can merely be caused by a sample consisting of individuals with low ability in the measured construct, and thus item and person parameters cannot be discerned within the framework of classical test theory. This is not the case in the Rasch modeling approach, a conceptual framework that belongs to the item response theory family. We therefore opine that an important scientific challenge is to explicitly test the underlying measurement model by means of item response theory analysis. Although, testing and verifying the latent measurement model is unfortunately frequently neglected in many domains of the social and health sciences, it is the indispensable basis for interpreting a scale score as well as achieving conceptual and psychometric invariance across samples and instruments (Bond and Fox 2007).

Which Scale Should Be Used?

The important question what scale is to be recommended cannot be generically answered but is rather dependent on the research context and purpose. One should, however, not forget that it is also reinforced by empirical support. Having reviewed the literature, we feel therefore confident to make some suggestions.

If the aim is to capture state-mindfulness, the TMS is the only instrument available for the time being. For assessing trait-mindfulness on the other hand, things are not so straight forward as several instruments are available. The predominantly used—and thus validated—scales are the KIMS and the MAAS. Despite recent criticisms (Grossman 2008, 2011), we would still recommend the MAAS as it has been proven to be useful in a plethora of studies. Utilizing this scale is presumably appropriate if a relatively short scale is needed and if the researcher is predominantly interested in a general mindfulness score rather than discerning the more fine-grained facets of mindfulness.

If it is important that the scale is aligned with Buddhist psychology as a theoretical ground of mindfulness, the FMI can be recommended. It has not been employed as frequently as the MAAS, but has seen a number of psychometric and experimental validation studies. For more fine-grained analyses, including the possibility to investigate different facets of mindfulness, the FFMQ can be recommended as this instrument not only incorporates other existing instruments (particularly the KIMS) but also differentiates several subcomponents of mindfulness. In cases in which not only mindfulness but also spiritual and meditative practice and behavior are of interest, the EOM scale may be considered. However, it should be noted that validation studies are scarce. Although the DMS is certainly innovative in its approach to provide internal construct validity by utilizing item response theory, we cannot explicitly recommend the DMS due to the lack of external validation at the moment. However, as relatively novel instruments (such as the MMS or the SMQ) cannot present the same amount of studies speaking in favor of their psychometric properties, it would be premature to dismiss such scales for the time being. Instead, we would recommend researchers to study the suitability of recently developed scales according to their own interests and research questions.

Alternatives to Questionnaire Instruments

Although the use of self-report psychometric questionnaires for measuring mindfulness is generally supported by this review, all questionnaires are beset with shortcomings that call for alternative ways to measure mindfulness. We would like to highlight some examples of alternative approaches.

Language-Based Measures and Qualitative Approaches

With regard to language-based measurement techniques, initial efforts have been made to demonstrate their appropriateness. The main advantage of this is that it is able to preserve semantically rich descriptions and in-depth details of the circumstances in which a measure is captured. For instance, Collins and colleagues (Collins et al. 2009) developed a language-based measure of mindfulness that takes the number of words related to mindfulness into account. The authors presented evidence that this language-based measure, derived from interviews with individuals participating in a mindfulness-based relapse prevention program, had predictive value as the number of words associated with mindfulness used by the participants was associated with less substance abuse in the 4 months following the treatment. This study provides initial support for the usefulness of interview-based methods in the context of mindfulness research that can also be combined with quantitative approaches. However, although this approach seems promising, it does not escape the principal criticism towards self-report questionnaires, as brought forward by Grossman (Oort et al. 2009). The main disadvantage remains: The individuals’ self-attributed mindfulness levels may differ from their “true” ability to be mindful; a shift in the internal reference point may occur as individuals may assess themselves as being more (or less) mindful after a mindfulness-based intervention. Furthermore, it should be noted that the approach of Collins et al (2009) presupposes that participants are sufficiently familiar with the conceptual framework of mindfulness.

Qualitative interview data can be valuable sources of information in its own right for three reasons. First and foremost, qualitative data allow for a more fine-grained investigation of subjective phenomena than quantitative approaches. Using interview data may, in some cases, be the better choice, for example, if the research interest is focused on intricate questions such as: Is the respondent understanding the question? Does the person have a similar understanding of mindfulness, so that results can be compared? Could a response shift potentially be involved? (Creswell 2003). Secondly, qualitative interview data may be used to identify different types of mindfulness as a more in-depth analysis is possible compared to quantitative approaches. Thirdly, interview data may actually complement quantitative approaches as their essence is explorative rather than confirmatory. A mixed-methods approach combining the advantages of quantitative and qualitative research can therefore provide insights that have not been preconceived earlier (Finucane and Mercer 2006; Kohls et al. 2008). Given that mindfulness may be embedded in a complex network of mutual interacting processes, such as motivational patterns, purposes pursued, personality traits, type of interventions maintained, and given that even different types of mindfulness may be extant, we would strongly recommend a mixed-method research (Creswell 2003).

In sum, explorative approaches are indispensable in order to increase the ability to answer complex research questions and to overcome simplistic cause–effect models. For sure, simple models are indispensable for most progress in most research areas, but at some point in the research process they frequently have to be replaced or at least complemented by research approaches that are thought to deepen and to further our qualitative understanding of the phenomena under investigation. The study of Finucane and Mercer (2006) is a constructive example for the usefulness of utilizing interview data for answering more complex research questions. The authors employed semi-structured interviews to explore the effects and benefits of an MBSR course. They used open questions such as “what aspects of the course did you find beneficial (difficult/unhelpful)?” As a result, insights into the perceived effects of specific parts of the mindfulness program, such as the “body scan” exercise could be obtained. Also, Lynch et al. (2011) used semi-structured interviews to gain insight as to how participants of a new mindfulness-based intervention for students perceive the training and how they employed and anchored mindfulness in their daily lives. Similar approaches were chosen for different populations such as rehabilitative contexts (Griffiths et al. 2009), chronic pain (Morone et al. 2008), Parkinson’s Disease (Fitzpatrick et al. 2010), and psychosis (Abba et al. 2008). Grounded theory approaches may also provide a fruitful avenue for deepening the understanding of mindfulness (Mason and Hargreaves 2001).

Qualitative studies seem to provide evidence for the two- or even multidimensional facets of mindfulness, as respondents are clearly able to differentiate between separate distinct components of their mindfulness experience(s). For example, Griffiths et al. (2009) found that their participants distinguished awareness and acceptance aspects of mindfulness. Interestingly, social and interpersonal aspects also seem to play a more pivotal role in interview data—an aspect underrepresented by concurrent psychometric scales. In contrast, psychometric questionnaire instruments tend to conceive mindfulness as an intrapersonal rather than interpersonal construct (Offenbächer et al. 2011). In sum, explorative approaches may be essential for shedding more light on the phenomenological aspects of mindfulness.

Assessment by Others

Assessment by others is an approach that is based on the external appraisal of the ability to be mindful. This procedure can only be applied if an external person is able to validly gauge the person’s mindfulness level or progress. For example, real-life diagnostics in sports are often based on the trainer’s opinion, without taking the player’s view into account. This “third person” approach is widely used and acknowledged as an assessment tool in practical contexts such as sports, teacher training, and performance judgment in business. External assessment from mindfulness educator reports may therefore constitute a potential avenue for measuring mindfulness. This approach is commonly employed in Buddhist contexts. It is based on the assumption that an expert trainer/teacher/educator is able to identify signs of progress or achievement on the basis of his or her own experiences. For instance, students may start to react with equanimity in a situation in which they used to behave short-temperedly. This or other subtle signs may be recognized by the trainer and thus provide the phenomenal criteria relevant for assessing change in the ability to be mindful. A drawback of the external assessment methods is that another person may be biased by his or her personal inclinations towards (or against) a certain individual. On the other hand, this approach is not likely to suffer from shifting baseline bias as the internal reference point of the trainer is not likely to change substantially during the course of the intervention. Also, experts such as trainers seem to be better equipped to understand mindfulness questionnaires and react more sensitively to them (Van Dam et al. 2009). To our knowledge, no study employing this approach has been published to date.

Biological and Neuropsychological Measures

A number of biological and neuropsychological correlates of mindfulness have been reported. There exists a plethora of electroencephalogram and brain scan studies investigating the neuronal correlates of mindfulness. For example, evidence exists that the cortical thickness in brain areas which were found to be active during states of mindfulness is larger in mindfulness practitioners compared to non-practitioners (Ott et al. 2011). Apart from the desirable properties such correlates may have, it is doubtful whether they can be termed as measurement of individual (mindfulness) levels. They may well prove to be sensitive to mindfulness interventions, but it is inappropriate to equalize the construct under investigation with its correlates (Michell 2005). Still, such approaches may be useful to differentiate between different levels of mindfulness on a group level. For example, Hölzel et al. (2011, 2008) found increased gray matter concentration in mindfulness meditators in certain cortical areas such as the right anterior insula. Furthermore, saliva cortisol levels have been shown to be sensitive to increase in mindfulness on group level, but so far not for individual assessment (Carlson et al. 2004; Lynch et al. 2011; Matousek et al. 2010; Matousek et al. 2011; Sudsuang et al. 1991).

Cognitive approaches, which could constitute another avenue for measuring mindfulness, are still in their infancy. Atmanspacher and Fach (2005), for example, speculate that the flicker rate in the perception of bistable images such as the Necker cube, may provide a base for measuring mindfulness. In the same vein, Sauer et al. (2012) have recently shown that individuals with mindfulness experience are able to stabilize one category of the Necker cube significantly longer if they are asked to enter a mindful state than individuals with lack of mindfulness training, suggesting that mindfulness may indeed influence the perception of bistable images. Similarly, Carter et al. (2005) showed that mindfulness experts such as Buddhist monks experience prolonged stabilization in binocular rivalry up to some minutes. Hence, neuropsychological measures, such as fluctuations of conscious states seem to be associated with perceptual rivalry. Standard attention tests, such as the Attention Network Test may prove to be promising candidates for objective testing and have already been used (Abba et al. 2008; Fan et al. 2005; Jha et al. 2007; Tang and Posner 2009; Zeidan et al. 2010). Other cognitive measures of mindfulness may exist, although so far none has been published.

Summary and Conclusion

In this article, we reviewed commonly used measurement approaches of mindfulness. With the field of research in mindfulness rapidly growing, a monomethod bias pointing to the usage of questionnaire instruments is noticeable. This is reflected by the fact that the amount of scales designed to measure mindfulness has proliferated. As it is important to compare different avenues for measuring mindfulness, this article provides an overview of the existing approaches and discusses strengths and weaknesses of different methods and tools in some detail. We conclude that although extant questionnaires for assessing mindfulness are, with regard to content and quality, partially comparable from a conceptual and psychometric point of view, all of them are in need of improvement. Having reviewed the scale and their applicability and frequency of use, it appears that while the psychometric approach reveals weaknesses and unsolved problems, it remains the most used and most practicable approach for measuring mindfulness for the time being. A conceptual consensus may be emerging suggesting that mindfulness is not unidimensional but may actually consist of two interrelated facets, present-moment attention and acceptance to the experience. We think that it is prudent to expand this line of research.

With regard to the question, “What questionnaire should be applied?,” we restate that there are no universal answers. It is rather the case that the optimal selection is largely dependent upon the research questions and objectives of the researcher. Nevertheless, the MAAS may be used if a short and straightforward instrument is needed. It should be noted, however, that there is mounting criticism about this scale. For a more in-depth analysis, we recommend taking a closer look at the FFMQ. The FMI may be useful if a tool in line with Buddhist theory is envisaged. Finally, for measuring mindfulness as a state, and not as a trait, as the majority of the scale do, the TMS is the only available option for the time being.

Other approaches have been utilized to a lesser degree. Promising candidates are the language-based measuring approach of Collins and colleagues (2009) and the research building on Grounded Theory, e.g., by Abba and colleagues (2008). As even the substantial number of psychometric scales have not yielded a consensus regarding the dimensionality of mindfulness, it may be advisable to better include insights from qualitative or mixed methods studies that have the ability to trace and embrace novel facets of mindfulness (Griffiths et al. 2009). However, these and other alternative measurement approaches potentially suffer from similar drawbacks than self-report questionnaires (e.g., different semantic constructions of mindfulness, social desirability). Approaches based on neuronal and neuropsychological parameters also appear to be able to distinguish between mindfulness practitioners and non-practitioners. However, for the time being, such methods are still not finely tuned enough to provide a valid diagnostic tool on an individual basis.

Finally, we would like to make some recommendations for improving the state of measurement of mindfulness. We recommend that researchers in the field of mindfulness measurement should strive to develop measurement models for the scales using item response theory and elaborate on qualitative methods such as the language-based measurement or mixed-methods approaches. Taking the criticism brought forward by Grossman and Van Dam (2011) seriously, testing for differential item understanding should become a mandatory obligation. For example, it should be tested whether mindfulness experts and novices have an equal understanding of the items under investigation. Elaborated methodology has been developed for that purpose (Austin et al. 2006; Choi et al. 2011). Additionally, although operationalizations of mindfulness seem to circle around the two factors presence and acceptance, questions regarding the taxonomy of the underlying facets of mindfulness remain to be answered. For the time being, we discourage the sole diagnostic usage of biological and neurological approaches for assessing mindfulness, as these methods have not yet provided evidence of sufficient specificity. However, cognitive and neuroscientific methods will without a doubt be important and necessary components to further the understanding of underlying processes and eventually measure mindfulness as “objectively” as possible. We should thereby not forget that the very essence of mindfulness will always be associated with subjective phenomenal states that cannot be reduced to their neuronal, psychological, and physiological correlates. However, we are confident that the combination of endeavors of both cognitive and neurological science, combined with methodological rigor and also creativity for developing new research avenues, will pave the way for the advance of the measurement of mindfulness.