Using Photographs in Mixed Methods Research: An Illustration Integrating Quantitative and Qualitative Analyses from Consumer Research

Abstract

Photograph analysis poses a novel methodological challenge for mixed methods researchers. In this paper, we argue that photographs are a valid data source that are not outside of the quantitative–qualitative binary and, hence, can be analyzed and used for integration, applying mixed methods principles. We summarize photograph analysis methods from different scientific fields and contribute to the field of mixed methods by proposing a mixed methods framework for analyzing visual data that allows the flexible application and integration of different quantitative and qualitative photograph analysis methods by focusing on data transformation. We use an illustrative example from consumer research.

Keywords

mixed methods research visual methods method development pictures

Introduction

Integration in mixed methods research (MMR) means that quantitative and qualitative data are combined at one or more stages in the research process in such a way that they add up to more than the sum of their parts (Woolley, 2009). Greene et al. (1989) and Fielding (2012) identified justifications for MMR, such as triangulation, complementarity, development, initiation, analytical density, and expansion. The choice of data type and the timing and method of the integration of qualitative and quantitative approaches are key (e.g., Fetters et al., 2013; Fielding, 2012; Tashakkori & Teddlie, 2007, 2010). Fetters et al. (2013) differentiates integration at the study design level (e.g., exploratory sequential, explanatory sequential, and convergent), at methods level (e.g., merging two databases for analysis), and at interpretation and reporting level (e.g., data transformation). The present article focuses on integration within the data analysis and interpretation level. Studies have adopted many different data integration strategies, including data transformation, typology, the consolidation of findings, case analyses, and building matrices (Bazeley, 2009; Caracelli & Greene, 1993). Data transformation, which is the focal interest in this article, means that one type of data is converted into the other type of data (e.g., qualitizing quantitative data or quantitizing qualitative data) so that the transformed data can be integrated with other data sets. For example, qualitative data can be transformed into numeric data and variables using content analysis and creating quantitative variables (e.g., Vogl, 2019), or quantitative data can be transformed into a qualitative format by, for instance, grouping participants based on quantitative measures (e.g., van Velzen, 2018).

The use of visual methods in MMR is highly varied and not new. However, few studies have systematically reflected on the use of visuals for integration. Plano Clark and Sanders (2015) found that visual displays in MMR have been used to depict different types of integration information in the form of matrices, figures, and tables. Others have added visual methods to complement, confirm, elaborate on, or contradict data generated with other methods (O’Connell, 2013) or have investigated the effect of visual activities (e.g., collage creation) on participants without analyzing the produced visuals (Johnson et al., 2020). There are examples of integrating quantitative or qualitative photograph analyses with non-visual methods, for example, Shannon et al. (2021) (integration of quantitative photograph analysis with qualitative group discussions) and Ares et al. (2011) (integration of semiotic qualitative photograph analysis with a quantitative survey); however, they do not integrate quantitative and qualitative photograph analysis.

Shannon-Baker and Edwards (2018) investigated the integration of visuals into mixed methods studies, positioning visual methods outside the qualitative-quantitative binary. Although photographs or other visual data may have unique characteristics, we think that such a juxtaposition of visual methods and mixed methods is not useful for a theoretical grounding of visual methods in MMR. We appeal instead for consideration of visual methods as one of many (qualitative and quantitative) methods that can be used and integrated in MMR, within the qualitative-quantitative paradigm. Indeed, as shown later using an example of analyzing photographs, visual analysis methods can apply either quantitative or qualitative principles (or both).

Conceptual Framework: Photographs and Visual Methods

Photographs have become an important part of society due to developments in digital photographic technology, the emergence of smartphones with integrated cameras and large storage capacities, and social media. Products and services are typically first experienced visually by consumers (e.g., on websites, at the point of sale, or in advertisements). Personal experiences with products or services are often shared through photographs. In short, the social habit of producing, sharing, and looking at photographs has grown in importance in recent years (Kleih et al., 2020).

Scholars increasingly employ photographs in empirical consumer research. Photographs have been used to, for example, investigate consumer preferences and consumption processes regarding clothes (e.g., Vermaak & Klerk, 2017), food (e.g., Thompson et al., 2015), and tourism (e.g., Giglio et al., 2019). However, most research approaches have fallen at one of two ends of a spectrum. On the one end, some scholars have used photographs merely as supportive stimuli for interviews and disregarded them in the analytical process (e.g., Vermaak & Klerk, 2017). On the other end of the spectrum are studies that have focused on the quantitative aspects of photos (e.g., Vilnai-Yavetz & Tifferet, 2015). More broadly, visual methods in the social sciences are in a difficult position, even though interest in the visual is high; a common theoretical grounding is still missing (e.g., Pauwels, 2010). Depending on the discipline, the significance given to visual methods differs. For example, sociology has had longstanding reservations about the validity and representativeness of visual materials, giving superiority to analysis of text or statistics, whereas in ethnography, there is a long tradition of the use of visuals (Grady, 2008; Pink, 2013).

As far as the term “visual data” is concerned, we adopt the approach used by Grady (2008) and use it to refer to any visual object or artifact, including material culture, images, pictures, and photographs. While we focus on photographs herein, we believe much of what we say can be transferred to other visual data. It will be useful for our discussion to differentiate between methods that analyze the figurative and semantic content of photographs (the subject of this article) and other visual methods, such as using photographs for documentation, conducting interviews, arts-based approaches, etc. (not the subjects of this article). By “photographic analysis,” we mean the systematic analysis of the content/internal narrative of what is displayed or represented in the photograph. However, we do not mean to suggest that a photograph has meaning based on the photograph in itself/ontologically, but rather that a subject (researcher/participant) attributes meaning to a photograph in an interpretative act. Approaches analyzing the content of photographs, such as in the qualitative social sciences, were developed for a detailed analysis of individual photographs. Nowadays, photographs are used quantitatively, as on websites or social media, meaning that such methods must be used in conjunction with other approaches (Kleih et al., 2020).

Methodological Aim

The methodological aim of this article is to depict ways to integrate quantitative and qualitative photograph analysis methods in the theoretical terms of MMR. To do so, we analyze the principles of existing photograph analysis methods and suggest a procedural framework for integrating these. We argue for the advantages of mixed methods photograph analyses. We also consider the integration of photograph data with other qualitative data (e.g., text obtained from qualitative interviews, comments on websites, or social media), as well as quantitative survey data. Neither quantitative nor qualitative photograph analysis is new, and neither is the combination of both (e.g., Bell & Milic, 2002); however, we think a thorough discussion of implementing these in MMR is missing and would be beneficial for the practice of MMR.

The Use and Challenges of Photographs in Social Science and Consumer Research

In the following, we first discuss the challenges of using photographs in social science research in general; and second, as we are coming from a consumer research point of view, we explore how photographs have been used in consumer research to date.

The Challenges of Using Photographs

Grady (2008) finds that some social science disciplines have reservations about visual data, for example, about how representative and valid such materials can be. The reasons for photographs’ undetermined role in some disciplines may be related to the challenges they pose for researchers due to their characteristics. In particular:

• Science is often interested in generalizations. The display of abstract facts is linked with non-iconic signs, whereas photographs are iconic and display concrete things (Wuggenig, 1989).The conclusion from concrete to general applies to analysis of other data types as well. So far, photographic analyses have met this challenge by displaying the general hypothesis derived from the analysis of concrete photographs with non-iconic signs (words).

• What photographs display is, on the one hand, obvious (e.g., a photo of a car shows a car) and, on the other hand, ambiguous (e.g., dependent on the viewer, cultural context, and interpretation process, a photo of a car may signify a prestige object or a means of transport). Furthermore, it is contested whether photographs have grammar (e.g., Forceville, 1999); everything is simultaneous. This polysemy and simultaneity make photographs difficult to use as scientific data (Fuhs, 1997). There are also complexities related to if, how, and when photographs should be translated into language (i.e., whether a transformation from iconic or presentative symbolization to discursive symbolization occurs), as well as questions related to whether photographs can be fully interpreted at all (see, e.g., Breckner, 2012; Müller-Doohm, 1997; Shannon-Baker & Edwards, 2018). To tackle these challenges, different photograph analysis methods have been developed (see Table 2). For example, the visual segment analysis (Breckner, 2012) draws on different ways of verbalization (e.g., physical, affective first impressions, and formal description) and analytical steps to articulate presentative symbolization of photographs.

• Ethical questions concerning the privacy and anonymization of research participants apply more strongly to photographs (O’Connell, 2013). For example, photographs may show participants’ faces, their private homes, or their social fields. The protection of such sensitive data is paramount. Researchers have addressed this issue either by cropping, blurring and excluding (parts of) photographs, or by asking participants at the end of data collection for their consent, so that they can agree to the usage and publication of specific visual content (e.g., Wills et al., 2013).

• Representation and power imbalance: The written word might be regarded as a superior medium of representation. Pink (2013) argues that there is no essential hierarchy of media. As with all other media, questions arise about how to represent participants and their experiences adequately (Banks & Zeitlyn, 2015). This is not just an epistemological matter, but also a matter of power (Galman, 2009). It directly concerns the analysis of photographs. The viewer of a photograph shares the experience of engagement with the photographer, but not necessarily the meaning (Grady, 2008). Hence, a researcher who performs an analysis should be aware of and make transparent the personal interpretative act, which is influenced by its disciplinary framing and other contexts. The result is a partial representation of the reality the researcher (in collaboration with the participant) perceived. The inherent power imbalance of representation has been addressed by replacing one voice (the researcher) with many, for example, through collaboration; shared reflection; and acknowledgment of the situatedness and cultural, historical, and contextual dependency of phenomena and constructedness of reality (Galman, 2009).

Visual methods encompass all sorts of involvement of visual data or visual work in the research process, not just the analysis of visual material. As we will see from the example of consumer research used below, the subject matter of research is often not the visual itself, but the reflective process or the effects of using visual methods.

The Use of Photographs in Consumer Research

In consumer research, using, and especially analyzing, photographs is still not very common, yet photographs are well-suited to consumer research. Consumer behavior is often studied for the purposes of improving existing or inventing new products and services (Naderer & Balzer, 2011). Thus, the research focus is usually a physical object or a service that has, in one way or another, a visual expression. According to sociology, physical objects do not only have instrumental functions; they also embody peoples’ emotions, personalities or lifestyles, and social statuses or goals (Wuggenig, 1989). Thus, photographs are capable of representing the material sides of consumers’ lives as well as the invisible values and meanings inherent in them. Just one viewing of a photograph immediately gives the viewer a sense not only of the content of the photograph but also of the atmosphere, context, and mood, which a mere written or verbal description would not be able to provide. In the following two paragraphs, we look at how photographs are used in qualitative and quantitative consumer research.

In qualitative consumer research, photographs are often used to capture consumers’ subjective perspectives and interpretations. Such methods include photovoice (e.g., Bibeau et al., 2012; O’Connell, 2013), photo-elicitation (e.g., Galvez et al., 2017), and photograph sorting (e.g., Lobinger & Brantner, 2020), which involves the integration of photographs into qualitative interviews. Photographs, which may be produced by either the researcher or the participant, are used during the interview process to activate memories, enhance communication and motivation, reduce stress, and provoke narratives (Collier, 1957). During the analysis, the main focus is on the transcribed interview text; the photographs themselves are not further analyzed (e.g., Bibeau et al., 2012; Galvez et al., 2017). In such cases, photographs are used to collect data, but are not viewed as data to be analyzed. The researchers’ interest lies not in what the photographs express, but rather on the participant’s reflection process and meaning making.

Quantitative methods and photographs have been used, for example, in e-commerce and advertising research (e.g., Taecharungroj & Mathayomchan, 2020), usability research, user-friendliness and consumer experience research (e.g., Vilnai-Yavetz & Tifferet, 2015), social media analysis (Tsai et al., 2016), and visual marketing (Kleih & Sparke, 2021). Photographs are sourced from consumers (e.g., Facebook and Instagram postings, and Flickr), the Internet, or existing advertising, and are analyzed using different software applications and deep-learning algorithms such as Google Cloud Vision API (e.g., Taecharungroj & Mathayomchan, 2020). Again, researchers consider the content of the photographs used to be less important than their effects on the consumers. The photographs’ obvious figurative meanings (e.g., the presence of certain objects, such as clothing) are also identified and quantified. Table 1 summarizes the research goals of the cited consumer research studies and how each uses and analyzes photographs.

In sum, consumer research involving photographs has either used them to support the analysis and investigation of other topics (mostly qualitative methods) or has analyzed their content on a descriptive level (mostly quantitative methods). Photographs have not been used for integration from an MMR perspective. Thus, we believe that consumer research would benefit from the integration of qualitative and quantitative visual analysis methods.

Table 1.

Research Goals and Usage of Photographs in Consumer Research Studies.

Source	Research goal	What kind of photographs are used?	How are photographs analyzed?
Bibeau et al., (2012)	To understand the perceptions of African American adults and teen girls regarding targeted food marketing.	Participant created photographs.	Photographs discussed in interview.
Galvez et al., (2017)	To explore the eating behaviors of Chilean women of low socioeconomic status.	Participants created photographs that represented what they considered important to them in their “food world”.	Photographs discussed in interview.
Giglio et al., (2019)	To understand consumers’ perception of luxury hotel brands.	Consumers’ pictures of hotel experiences.	Quantitative analysis of consumer pictures by machine learning approach.
Kleih and Sparke, (2021)	To investigate the importance and consumer recognition of fruit brands at fruit displays in supermarkets.	Researcher produced photographs of fresh fruit and vegetable supermarket displays.	The perception of consumers via eye tracking technology was quantitatively analyzed, as well as discussed in interview.
Lobinger and Brantner, (2020)	To describe card-sorting in general and to elaborate on why this research technique is particularly suited for visual research.	Cards/pictures.	Analysis of participants’ sorting and ranking of pictures.
O’Connell, (2013)	To explore the usefulness of using visual methods to study children’s food practices.	Children were to, among others, photograph about their food practices.	Data from VM was compared to interview data.
Taecharungroj and Mathayomchan, (2020)	To identify and analzye city image dimension.	Photographs of cities from Flickr.com	Quantitative analysis of consumer pictures by machine learning approach.
Thompson et al., (2015)	To examine what it means to be an adult picky eater.	Participants were to take pictures of everything they ate and drank, where and with whom for 4-day period.	Photographs discussed in interview.
Tsai et al., (2016)	To estimate the time of capture in consumer photos.	Consumer photographs in travelling; photographs from a photo sharing website; live online cameras.	Photographs were analyzed with a novel system, based on astronomy, to determine time of capture.
Vermaak and Klerk, (2017)	To explore and describe young millennial clothing consumers' dressing room experiences.	Participants produced photographs of retailers' dressing rooms before the second interview.	Photographs served as confirmation of the verbatim.
Vilnai-Yavetz and Tifferet, (2015)	To segmenting consumers by Facebook profile images.	Consumer segmentation of Facebook users based on the way they use profile picture.	Profile images were quantitatively coded.

Photograph Analysis Methods

Overview of Photograph Analysis Methods

In the following, we give a brief overview of photograph analysis methods (which may also be applied to other visual materials, such as art) from the qualitative social sciences, the science of art, and visual communication science in order to familiarize the reader with existing analysis strategies. Table 2 gives a summary of the analysis methods, their approach to visual data, and how visual data is transformed during the analysis. The different analysis methods have different understandings of how photographs convey meaning and how meaning is constructed. We will present one qualitative (visual segment analysis) and one quantitative (visual content analysis) approach in more detail, as we will use both for our empiric example to illustrate our suggested integration framework.

Table 2.

Overview of Photograph Analysis Methods.

Photograph analysis method	Meaning generation/construction	Transformation of the visual data during analysis into
Documentary method (Bohnsack, 2006)	- Formal composition/structure of the photograph. Homology between formal and content-wise meaning structure - At first bracketing/suppressing of textual-narrative pre-knowledge	- Words/Text
Objective hermeneutics (Peez, 2004)	- Meaning is constituted through process and sequentially/chronologically	- Words/Text
Visual segment analysis (Breckner, 2012)	- Meaning originates from formal elements and in the act of perception; specific internal and external meaning relations: indexical, iconic, and knowledge-based references	- Words/Text
Semiology/semiotic (Barthes, 1981; Peirce, 1965; Saussure, 2003; Yakin & Totu, 2014)	- Meaning arises from certain forms of visual presentation (denotation) and due to cultural knowledge/associations (connotation) - Duality of signifier and signified - Trichotomy of sign/representatum, object/referent, interpretant	- Words/Text
Structural-hermeneutical symbol-analysis (Müller-Doohm, 1997)	- Visual order and structure of picture elements structures and guides meaning. The interpretation derives from cultural-sociological theories and concepts	- Words/Text
Figurative hermeneutic and picture cluster (Müller, 2012, 2016)	- Meaning arises through comparison with other photographs in a prelingual manner. How was a similar topic already displayed or how could one possibly be displayed?	- Words/Text
Grounded theory method/methodology for visual communication analysis (Kautt, 2017)	- The visible form of material objects must be joined by a situationally anchored practice of meaningful interpretation by people and by a linking of situational practice to structures	- Words/Text
Iconological-iconographical method (Panofsky, 1975)	- Based on context (literature, philosophy, historical thinking, and perceptions)	- Words/Text
Iconic (Imdahl, 1994)	- Immanent, inherent logic; based on internal structure of photograph; photograph itself is carrier of meaning	- Words/Text
Quantitative visual content analysis (e.g., Bell, 2010)	- Focus is not meaning of photographs, rather quantitative appearances of certain photograph motives in a bulk of photographs	- Numbers/Categories

Visual Segment Analysis ‐ Example of a Qualitative Approach

Breckner (2012) suggests a symbol-theoretical approach to analyze photographs. Its starting point is the idea that symbolization processes generate meaning. Photographs emerge from symbolization processes and they connect to previous symbolization processes. By looking at photographs, visual perception processes emerge, which have photograph-specific dynamics that cannot be articulated verbally straightforwardly. During verbalization, photographs and their meaning change. The different modes of generating meaning between photographs and language/text are as follows: the meaning-generating elements (Breckner calls them “segments”) in photographs have no syntax and their perception follows no linear path. Visual perception takes place simultaneously and successively. While all elements are present at the same time, perception is a process, and focuses on different elements one by one and puts them into relation. Through this process, meaning is generated. The process of putting elements in relation is based on indexical, knowledge based, and iconic references in the photograph. The processes of perception, concept formation, and interpretation are closely and insolubly connected with each other. In other words, the meaning of a photograph and its perception are guided, but not determined, by the iconic structure of the photograph.

Quantitative Visual Content Analysis ‐ Example of a Quantitative Approach

Visual content analysis is a systematic method that examines visual patterns by categorizing and quantifying units of content, rather than examining individual photographs. A visual content analysis can identify which content is treated as important and which agenda is followed with which visual strategy. The research questions of such analyses focus on the priority given to certain content (e.g., how often, how large, or where in sequence is something shown), the frequency of certain topics, and historical changes in representations (Bell, 2010). Visual content analyses now frequently use computer deep-learning technology; for example, commercial picture analysis software and artificial intelligence can identify and count or measure objects and have been used to analyze visual content, particularly in social media (Giglio et al., 2019; Zhang et al., 2019). The analytical process requires, first, the definition of the visual content and the visual body to be analyzed (i.e., which visual unit or topic is in focus and in which media?) Second, the specified visual content is identified within the whole photograph body and is analyzed (e.g., by counting).

Combination of Quantitative Content Analysis and Visual Semiotic

Some studies have used combined approaches (e.g., Bell & Milic, 2002) to categorize and interpret a large volume of photographs. Bell and Milic (2002), for example, isolated semiotic variables and used them quantitatively as codes. Mortelmans (1997), proceeding in a similar manner, analyzed advertisements by decomposing advertisement’s signifier and signified (sign) into a form and a content each in order to apply this coding quantitatively. His analysis did not aim to construct the meanings of individual advertisements, but to identify the constructive process that results in a luxury signified.

Reflection on Photograph Analysis Methods

The qualitative approaches described in the section “photograph analysis methods” (see Table 2 and the section “visual segment analysis”) analyze in detail the composition of a photograph and investigate, either implicitly or explicitly, the question of how photographs convey meaning or how meaning can be attributed to the photograph. These methods differentiate between two levels of meaning: the obvious meaning that is depicted in the photograph (i.e., the immanent meaning, denotation, or manifest meaning) and the more subtle meaning that requires interpretation and reconstruction (i.e., the documentary meaning, connotation, or latent meaning). The qualitative approaches consider that photographs and their interpretation construct reality, whereas the quantitative approach captures their manifest meaning (e.g., the appearance of certain topics or their patterns in the majority of photographs) and treats them as real in and of themselves. Therefore, an integration of qualitative and quantitative analysis methods needs to answer research questions that address both levels of meaning.

Studies that quantify visual semiotics show how these levels of meaning can be addressed by combining quantitative and qualitative methods. We adopt this notion for a generalized theoretical discussion of how quantitative and qualitative photograph analysis can be integrated. The qualitative approaches translate the photographs into text in various ways and the quantitative approach quantifies their themes, similarly to the quantification of text through coding (data transformation). The resulting data, the text, codes, and numbers, are the connecting points for the integration of these methods. As stated above, the notion of “translating” photographs into text (and quantities) is contested, as is the possibility of and even need for analyzing photographic content at all. We do not take the stance that this is the only way of handling visuals in social research. Instead, we argue in favor of the pragmatic approach that this process offers the possibility of integrating quantitative and qualitative photograph analysis.

Using Photographs for Integration in Mixed Methods Research

Integration of Quantitative and Qualitative Photograph Analyses

Integration in MMR means that quantitative and qualitative methods are combined in such a way that together they add up to more than the sum of their parts (Woolley, 2009). The transformation of photographs into text (qualitative) and numeric (quantitative) material, and the levels of meaning (i.e., manifest and latent) are the starting points for integration in our framework. Figure 1 depicts our proposed procedural framework for integrating quantitative and qualitative photograph analysis methods.

Figure 1.

Procedural framework for integrating quantitative and qualitative photograph analyses.

Our proposed approach is based on the assumption that a large number of photographs have been sampled as data to be analyzed. The analysis could start with a quantitative content analysis by defining categories and categorizing all of the photographs, possibly with the help of computer software. On basis of this categorization, frequencies, overlaps, and patterns can be statistically analyzed. The quantitative content analysis investigates the manifest meaning of the photographs on the metonymic level and forms the basis of the qualitative photograph analysis that follows. In the latter, individual photographs from each category can be selected based on specific qualities, such as being prototypes or representatives of a category, representing extreme cases, or showing contrast, or based on a typology. The selected photographs can be further analyzed using qualitative methods, investigating their latent meaning on the metaphorical level. The analysis could also start with a qualitative photograph analysis (e.g., identifying how certain connotations are constructed), then use this finding to guide the sampling of photographs by searching for and quantifying those elements (e.g., how often and in which context do they occur?). The results of the quantitative and qualitative analyses can be integrated through comparison, interpretation, explanation, and writing cases.

Empiric Example of the Procedural Framework in Consumer Research

To clarify the process depicted in Figure 1, we draft an empirical example. We take the data from a study in Germany, where we asked seven participants to take photographs of their consumer behavior regarding fresh fruits and vegetables. The participants were chosen based on their household situation (single, with partner, with children etc.), age, and gender. Figure 2 shows a selection from the 171 photographs we received. We printed the photographs and conducted an interview with each participant to talk about the photographs. For the analysis, we used an explanatory mixed methods research design. For the quantitative photograph analysis, we applied the quantitative visual content analysis, and for the qualitative photograph analysis, the visual segment analysis.

Figure 2.

Sample of participant photograph material.

Quantitative Photograph Data Analysis Phase of the Empiric Example

Category-building process: Based on the theory of practice (see Reckwitz, 2002; Warde, 2005) and drawing from Halkier (2009), who applied the theory of practice empirically, we used activities, understandings, procedures, and engagements as the basic category framework, which we applied to both the photographs and the interview transcript (deductive category building). During the quantitative photograph analysis, we looked through the photographs to identify different kinds of fruit and vegetable consumption activities (inductive category building). We identified the following categories, understood as activities (frequencies): shopping (20), preparing/handling (56), eating (45), gardening (13), playing and making art (12), and reflecting (19), as well as residuals (6). During coding, we realized that the photographs included the material site of practices, the technologies involved, and “‘things’ and their use” (Reckwitz, 2002, p. 249). To define the categories, we considered the following questions: What is typical about the category and what do photographs belonging to this category have in common (internal homogeneity)? What is the difference between this and other categories (external heterogeneity)? Frequencies of each category were added. To identify the most common activities, we counted how often each activity was depicted. The results of the quantitative photograph analysis helped us to cover and describe the whole photograph corpus, to identify different kinds of fruit and vegetable consumption activities (e.g., preparing/handling), and to see how their performances are materially displayed (e.g., for “chopping” knives, chopping boards, half cut vegetables and fruits, and food processors).

Integration During Analysis Phase of the Empiric Example

In order to integrate the qualitative photograph analysis with the quantitative analysis during the analysis process, we selected, based on the above-described definition of a category, one photograph to represent each category in a typical way. For example, we chose the photograph depicted in Figure 3 for the sub-category “chopping” within the category “preparing.”

Figure 3.

Selected photograph for qualitative analysis.

Qualitative Photograph Data Analysis Phase of the Empiric Example

The selected single photographs were analyzed using a process based on visual segment analysis (Breckner, 2012). The interpretation process involves three steps. The first step includes the immediate photo perception and articulation of first impressions (Figure 4). The second step is the formal description with the identification of segments (units of meaning that build up the photograph, Figure 5.

Figure 4.

Depiction of self-reflective perceptive process of interpreter (first step).

Figure 5.

Identified segments.

The third step is the description and interpretation of the single segments, their indexical, symbolical, iconic references and their relationship to the photograph surface and the combination of segments. This step can be enriched by (literature) research and applying different ways of seeing to corresponding possible photograph contexts. To illustrate the analysis process, excerpts of the interpretation during the third step for the segment “pears” are below:

Two halves of a pear can be seen. The formerly whole bulb has been cut, the cut-open side is facing upwards. This suggests that the pear is probably being prepared for consumption. Though often, a pear is consumed with its peel. Other contexts would be imaginable, such as a scientific investigation, checking for infestation, or displaying a pear for pedagogical purposes. Pears are fruits and commonly praised for their nutritional value. Thus, they fall under the social discourse of “health.” When comparing the pear with the apple, its advantage of having lower levels of acid is emphasized, which, in particular, makes it easier for babies to digest. Finally, the pear found its way into German poetry through the famous poem “Squire of Ribbeck at Ribbeck in Havelland.” This poem is taught in school and can be found in children’s books. It is about a man who gave pears to children until his death and then, by bringing a pear into his grave, made a pear tree grow to provide the fruits over the generations. The poem brings pears into the context of generosity and care (especially toward children) and of giving blessings even in death.

Summarizing the interpretations of the various segments, we conclude: the pear is a positively associated, healthy food, which is particularly suitable for children and popularly stands for generosity and care, and is aesthetically “presented” here in its preparation. The contrast between the white pear, the bright green cutting board, and the red knife handle and the careful alignment of the objects give a deliberately esthetic impression, even though the presumed action may be an everyday one. The mundane is arranged in a scene that suggests particular care, care presumably directed at a child, as can be inferred from the design of the cutting board, the small plastic spoon, and the manner of preparation (peeling). It may be concluded that the preparation of this pear is an act of care, love, and appreciation. Considering the strong association with fruits as a healthy food today, this photograph may also reveal what the photographer understands to be good parenting or how he or she understands the role of parenting.

Integration of the Quantitative and Qualitative Results Phase of the Empiric Example

The qualitative interpretation of a single photograph shows that it enriches the quantitative analysis and vice versa. The qualitative analysis gave the fruit and vegetable activities identified in the quantitative analysis a deeper dimension. It was possible to work out in which context a certain fruit and vegetable activity was carried out and according to which engagements, understandings, or procedures the respective activity was coordinated.

Empiric example of the Integration of photograph data analyses with other data analyses

The ways of how to integrate photograph data analyses with other data analyses are manifold. Shannon-Baker and Edwards (2018) suggested three ways of integrating visual data with other forms of data by drawing from three example studies: writing cases, quantifying visual images, and making cross-comparisons (e.g., data matrices). Transferred to our framework, we quantify the visual data during the quantitative content analysis. The resulting categories/variables could be integrated into other statistical data sets. Furthermore, the defined categories could also serve as deductive codes for qualitative text data, as follows:

The transcribed interviews were analyzed by deductive and inductive coding. We used the identified activities (e.g., preparing/handling) from the quantitative photograph analysis and results of the qualitative analysis (e.g., parenting) as codes (deductive), but also developed new codes from the text (inductive) to further capture the understandings, procedures, and engagements that structure fruit and vegetable consumption. We found dimensions such as “pleasure principle,” “time & practical constraints,” and “moral of healthiness,” many of which apply not only to fruit and vegetable consumption, but to food consumption or everyday practices in general. To integrate the text analysis with the photograph analysis, we compared the codes and the respective text segments in which the participant talked about the photograph, leading to a richer understanding.

The analyzed photograph came from a 31-year-old man, married, with a 9-month-old daughter. At the time of the interview, he was on parental leave, taking care of the baby and much of the household. His everyday life was structured according to the needs of his daughter. When he talked about the photograph, he became very particular, using a visual language:

“Yes, here, exactly, we had already addressed that I just, so chop a pear into small pieces. […] Compared to an apple or something like that, which is still much too hard when you only have two little teeth […]. And um, and grapes […], they have also a skin. I chew them, that is, I open them in the middle, bite out the pulp, and give it back to her. Like a mother bird and then eat the rest of it. Yes.”

During other parts of the interview, he noted that he recognized that he takes much more care about what his daughter eats than what he himself eats, especially with reference to fruits and vegetables. His mother used to cut fruits for him when he was a child; nowadays he and his wife only do it when one of them is sick. Hence, the cutting of fruits expresses caring and is bound into the practices of caring and parenting.

The benefits of integrating the photograph analysis with the interview analysis is that it adds up to a condensed understanding of the photographs. It includes the intended meaning of the participants and may offer new facets, add, or even contradict the photograph interpretation.

Discussion and contribution to the field of mixed methods research

There are different ways of integrating data in mixed methods analysis (Fetters et al., 2013). Most social science studies do not employ full data integration; instead, the analyses are held separately and the results are then compared (Bryman, 2006). By contrast, our framework suggests how quantitative and qualitative photograph data analyses can influence each other during the analytical process. We show how integration with visual material (photographs) is possible by using data transformation: photographic material can be transformed into quantities or qualitative interpretations. In this sense, we demonstrated that the integration of different kinds of data in MMR can be extended by photographic material as another source of data. It goes beyond visual displays (Plano Clark & Sanders, 2015), or using photographs parallel to other types of data (e.g., O’Connell, 2013). Our example of integrating quantitative and qualitative photograph analyses is however, consecutive. The transformed visual data is not integrated into another data set. Our example starts with the quantitative photograph analysis, which informs the consecutive qualitative photograph analysis. It is also conceivable to start with a qualitative photograph analysis, and to use the findings to guide the quantitative photograph analysis. Drawing from our example, we could, based on the qualitative analysis, look for other displays and uses of pears or for photographs from the same vantage point, etc.

Our example shows how quantitative photograph analysis can be integrated with qualitative text analysis by using the emerged categories from the quantitative photograph analysis as deductive codes (added by inductive coding). Integration with other quantitative data (e.g., variables from surveys) is possible by, for instance, merging the quantitative data set which resulted from quantitative photograph analysis with another quantitative data set.

The integration of qualitative and quantitative photograph analysis can be applied to different mixed methods designs with either quantitative or qualitative priority (or both), and be mixed with different data types. Our empiric example followed an explanatory mixed methods research design in the analysis (first quantitative, then qualitative based on quantitative); we also mentioned how an exploratory mixed methods research design is possible (first qualitative, then quantitative based on qualitative). While it was beyond the scope of our empiric example, our suggested procedural framework could be used in convergent mixed methods research designs (concurrent quantitative and qualitative data collection) as well.

A distinctiveness of our mixed methods photograph analysis framework is that both the quantitative and qualitative approaches use the same data. The sampling of photograph data can occur either quantitatively (e.g., using software to select photographs within predefined categories) or qualitatively (e.g., based on a qualitative interview and open research question), and the resulting material can be treated both qualitatively and quantitatively. Hence, we agree with Vogl (2019) that data are constructed and not inherently qualitative or quantitative, and it may be discussed if the data processing makes them qualitative or quantitative.

The principles for the integration of quantitative and qualitative photograph data analysis methods, as well as the rationale for integration and means of transformation, are similar to those of other mixed methods data analyses: The selection of single photographs for a qualitative analysis can be based on a typology, can identify prototypes of a category, or can be understood as an intensive case analysis or case study (see Bazeley, 2009). Therefore, the suggested framework is congruent with the four analytical strategies in MMR (Caracelli & Greene, 1993). In our empiric example, the integration of photographic data analyses (quantitative and qualitative) was key for identifying and understanding the fruit and vegetable consumption activities. In the quantitative analysis, the different activities and their material expressions were determined, while in the qualitative analysis, the motivation and engagement that structure these activities were identified. Hence, our suggested framework enables a more comprehensive analysis of photographic material, as both levels expand and complement each other in the interpretation process, which is in line with the purposes for mixing methods identified by Greene et al. (1989) (complementarity and expansion) and Fielding (2012) (analytical density).

Limitations and Opportunities for Future Research

A shortcoming of our study is that we did not conduct a systematic literature review of existing photograph analysis methods. Consequently, our overview of such methods (Table 2) is neither mutually exclusive nor collectively exhaustive. This decision reflects the vast variety of such approaches, which is beyond the scope of this paper (for a more detailed review, see, e.g., Knoblauch et al., 2008). A systematic review may be another subject matter for further research. Furthermore, in future, the suggested framework and its different application possibilities should be fleshed out with more empiric examples. Our purpose here was to show in principle what the integration of qualitative and quantitative photograph analysis could look like and how it could be theoretically discussed in terms of MMR. Additionally, we focused our study on the analytical process, and not on photograph data sampling or the reporting of results of integrated quantitative and qualitative photograph analyses. Finally, in our empiric example, the researchers interpreted the photographs void of participants’ voice. This approach has the drawback that the (production) context of the photograph and participants’ intention for making this photograph are not considered during this analysis step.

Conclusion

The prevalence of photography in society requires the inclusion of photographs in market research and social science and calls for a comprehensive analysis of such photographic material. In this paper, we provided an overview of different qualitative and quantitative photograph analysis methods from different scientific fields and found that, in each method, photographs are transformed into text or quantitative data in some way. We showed that the integration in MMR can be enhanced by photographic material, using the mixed methods principles of data transformation, consolidation, typology, and extreme case analysis. As quantitative and qualitative photograph analyses investigate different levels of meaning, a mixed methods approach has the advantages of expansion and complementarity. Due to the transformation of photograph data into text and numbers, integration with other data types is also possible. Our work considered the challenge of integrating qualitative and quantitative data in the context of the analysis of photographs and demonstrates that photographs as a data source do not exist outside the qualitative-quantitative paradigm and hence fit into the conceptualizations of MMR. Depending on how photographs are used, they can be used as either qualitative or quantitative data. We hope that our contribution encourages the use of photographs in the research process for integration and stimulates both future reflection and application of integrated quantitative and qualitative photograph analyses.

Footnotes

Acknowledgments

The work was conducted within a joint PhD program at the Department of Agricultural Sciences, Ecotrophology, and Environmental Management, Justus-Liebig University Giessen.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Anne-Katrin Kleih

References

Ares

Piqueras-Fiszman

Varela

Marco

R. M.

López

A. M.

Fiszman

(2011). Food labels: Do consumers perceive what semiotics want to convey? Food Quality and Preference, 22(7), 689–698. https://doi.org/10.1016/j.foodqual.2011.05.006

Banks

Zeitlyn

(2015). Visual methods in social research (2nd ed.). SAGE.

Barthes

(1981). Elemente der Semiologie (2nd ed.). Syndikat.

Bazeley

(2009). Editorial: Integrating data analyses in mixed methods research. Journal of Mixed Methods Research, 3(3), 203–207. https://doi.org/10.1177/1558689809334443

Bell

(2010). Content analysis of visual images. In van Leeuwen

Jewitt

(Eds.), Handbook of visual analysis (pp. 10–34). SAGE.

Bell

Milic

(2002). Goffman’s gender advertisements revisited: Combining content analysis with semiotic analysis. Visual Communication, 1(2), 203–222. https://doi.org/10.1177/147035720200100205

Bibeau

W. S.

Saksvig

B. I.

Gittelsohn

Williams

Jones

Young

D. R.

(2012). Perceptions of the food marketing environment among African American teen girls and adults. Appetite, 58(1), 396–399. https://doi.org/10.1016/j.appet.2011.11.004

Bohnsack

(2006). Die dokumentarische Methode der Bildinterpretation in der Forschungspraxis. In Marotzki

Niesyto

(Eds.), Medienbildung und Gesellschaft: Vol. 2. Bildinterpretation und Bildverstehen: Methodische Ansätze aus sozialwissenschaftlicher, kunst- und medienpädagogischer Perspektive (pp. 45–75). VS Verlag für Sozialwissenschaften | GWV Fachverlage GmbH Wiesbaden.

Breckner

(2012). Bildwahrnehmung – Bildinterpretation. Österreichische Zeitschrift Für Soziologie, 37(2), 143–164. https://doi.org/10.1007/s11614-012-0026-6

10.

Bryman

(2006). Integrating quantitative and qualitative research: How is it done? Qualitative Research, 6(1), 97–113. https://doi.org/10.1177/1468794106058877

11.

Caracelli

V. J.

Greene

J. C.

(1993). Data analysis strategies for mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 15(2), 195. https://doi.org/10.2307/1164421

12.

Collier

(1957). Photography in anthropology: A report on two experiments. American Anthropologist, 59(5), 843–859. https://doi.org/10.1525/aa.1957.59.5.02a00100

13.

Fetters

M. D.

Curry

L. A.

Creswell

J. W.

(2013). Achieving integration in mixed methods designs-principles and practices. Health Services Research, 48(6 Pt 2), 2134–2156. https://doi.org/10.1111/1475-6773.12117

14.

Fielding

N. G.

(2012). Triangulation and mixed methods designs. Journal of Mixed Methods Research, 6(2), 124–136. https://doi.org/10.1177/1558689812437101

15.

Forceville

(1999). Educating the eye? Kress and Van Leeuwen’s reading images: The grammar of visual design (1996). Language and Literature: International Journal of Stylistics, 8(2), 163–178. https://doi.org/10.1177/096394709900800204

16.

Fuhs

(1997). Fotografie und qualitative Forschung: Zur Verwendung fotografischer Quellen in den Erziehungswissenschaften. In Friebertshäuser

Prengel

(Eds.), Handbuch qualitative Forschungsmethoden in der Erziehungswissenschaft (pp. 265–285). Juventa Verlag. https://doi.org/10.1007/978-3-322-97489-1_3

17.

Galman

S. A.C.

(2009). The truthful messenger: Visual methods and representation in qualitative research in education. Qualitative Research, 9(2), 197–217. https://doi.org/10.1177/1468794108099321

18.

Galvez

P., E.

Vizcarra

Palomino

A. M.

Valencia

Iglesias

Schwingel

(2017). The photo-elicitation of food worlds: A study on the eating behaviors of low socioeconomic Chilean women. Appetite, 111, 96–104. https://doi.org/10.1016/j.appet.2016.12.040

19.

Giglio

Pantano

Bilotta

Melewar

T. C.

(2019). Branding luxury hotels: Evidence from the analysis of consumers’ “big” visual data on TripAdvisor. Journal of Business Research. Advance online publication. https://doi.org/10.1016/j.jbusres.2019.10.053

20.

Grady

(2008). Visual research at the crossroads. Forum Qualitative Sozialforschunbg/Forum Qualitative Social Research, 9(3), Art. 38. https://doi.org/10.17169/FQS-9.3.1173.

21.

Greene

J. C.

Caracelli

V. J.

Graham

W. F.

(1989). Toward a conceptual framework for mixed-method evaluation designs. Educational Evaluation and Policy Analysis, 11(3), 255–274. https://doi.org/10.3102/01623737011003255

22.

Halkier

(2009). A practice theoretical perspective on everyday dealings with environmental challenges of food consumption. Anthropology of Food. Advance online publication. https://doi.org/10.4000/aof.6405

23.

Imdahl

(1994). Ikonik: Bilder und ihre Anschauung. In Boehm

(Ed.), Was ist ein Bild? (pp. 300–324). Fink.

24.

Johnson

J. K.

Carpenter

Goodhart

Stewart

A. L.

Du Plessis

Coaston

Clark

Lazar

Chapline

(2020). Exploring the effects of visual and literary arts interventions on psychosocial well-being of diverse older adults: A mixed methods pilot study. Arts & Health, 13(3), 263–277. https://doi.org/10.1080/17533015.2020.1802603

25.

Kautt

(2017). Grounded theory als Methodologie und Methode der Analyse visueller Kommunikation. Forum Qualitative Sozialforschunbg/Forum Qualitative Social Research, 18(3), 8. http://doi.org/10.17169/fqs-18.3.2859

26.

Lobinger

Brantner

(2020). Picture-sorting techniques. Card sorting and Q-sort as alternative and complementary approaches in visual social research. In Pauwels

Mannay

(Eds.), The sage handbook of visual research methods (2nd ed., pp. 309–321). SAGE.

27.

Mortelmans

(1997). Visual represeratation of luxury. An analysis of print advertisements for jewelry. Communications, 22(1), 69–91. https://doi.org/10.1515/comm.1997.22.1.69

28.

Müller

M. R.

(2012). Figurative Hermeneutik: Zur methodologischen Konzeption einer Wissenssoziologie des Bildes. Sozialer Sinn, 13(1), 129–161. https://doi.org/10.1515/sosi-2012-0107

29.

Müller

M. R.

(2016). Bildcluster: Zur Hermeneutik einer veränderten sozialen Gebrauchsweise der Fotografie. Sozialer Sinn, 17(1). https://doi.org/10.1515/sosi-2016-0004

30.

Müller-Doohm

(1997). Bildinterpretation als struktural-hermeneutische Symbolanalyse. In Hitzler

Honer

(Eds.), Uni-Taschenbücher Sozialwissenschaften: Vol. 1885. Sozialwissenschaftliche Hermeneutik: Eine Einführung (pp. 81–108). Leske + Budrich.

31.

Naderer

Balzer

(Eds.). (2011). Qualitative Marktforschung in Theorie und Praxis (2nd ed.). Gabler Verlag. http://gbv.eblib.com/patron/FullRecord.aspx?p=750452

32.

O’Connell

(2013). The use of visual methods with children in a mixed methods study of family food practices. International Journal of Social Research Methodology, 16(1), 31–46. https://doi.org/10.1080/13645579.2011.647517

33.

Panofsky

(1975). Ikonographie und Ikonologie: Eine Einführung in die Kunst der Renaissance. In Panofsky

(Ed.), Sinn und Deutung in der Bildenden Kunst. Dumont Literatur U. Kunst.

34.

Pauwels

(2010). Visual sociology reframed: An analytical synthesis and discussion of visual methods in social and cultural research. Sociological Methods & Research, 38(4), 545–581. https://doi.org/10.1177/0049124110366233

35.

Peez

(2004). Im Foto ist alles gleichzeitig: Wie sich die Simultaneität von Fotografien und die Sequenzialität von schriftsprachlichen Texten in qualitativer Empirie aufeinander beziehen lassen. Medien Pädagogik, 9(Visuelle Methoden), 1–24. https://doi.org/10.21240/mpaed/09/2004.11.16.X

36.

Pink

(2013). Doing visual ethnography (3rd ed.). Sage Publ.

37.

Plano Clark

V. L.

Sanders

(2015). The use of visual displays in mixed methods research: Strategies for effectively integrating the quantitative and qualitative components of a study. In McCrudden

M. T.

Schraw

Buckendahl

McCrudden

(Eds.), Current perspectives on cognition, learning and instruction. use of visual displays in research and testing: coding, interpreting, and reporting data (pp. 177–206). Information Age Publishing.

38.

Reckwitz

(2002). Toward a theory of social practices: A development in culturalist theorizing. European Journal of Social Theory, 5(2), 243–263. https://doi.org/10.1177/13684310222225432

39.

Saussure

F. d.

1. ). (2003). Course in general lingusitics. In Gottdiener

Boklund-Lagopoulou

Lagopoulos

A. P.

(Eds.), Semiotics. Sage Publications.

40.

Shannon-Baker

Edwards

(2018). The affordances and challenges to incorporating visual methods in mixed methods research. American Behavioral Scientist, 62(7), 935–955. https://doi.org/10.1177/0002764218772671

41.

Shannon

Borron

Kurtz

Weaver

(2021). Re-envisioning emergency food systems using photovoice and concept mapping. Journal of Mixed Methods Research, 15(1), 114–137. https://doi.org/10.1177/1558689820933778

42.

Taecharungroj

Mathayomchan

(2020). The big picture of cities: Analysing Flickr photos of 222 cities worldwide. Cities, 102, 102741. https://doi.org/10.1016/j.cities.2020.102741

43.

Tashakkori

Teddlie

(Eds.). (2007). Handbook of mixed methods in social & behavioral research [Nachdr.]. Sage Publications. http://www.loc.gov/catdir/enhancements/fy0655/2002005206-d.html

44.

Tashakkori

Teddlie

(Eds.). (2010). Sage handbook of mixed methods in social & behavioral research (2nd edition). SAGE.

45.

Thompson

Cummins

Brown

Kyle

(2015). What does it mean to be a 'picky eater'? A qualitative study of food related identities and practices. Appetite, 84, 235–239. https://doi.org/10.1016/j.appet.2014.09.028

46.

Tsai

T.-H.

Jhou

W.-C.

Cheng

W.-H.

M.-C.

Shen

I.-C.

Lim

Hua

K.-L.

Ghoneim

Hossain

M. A.

Hidayati

S. C.

(2016). Photo sundial: Estimating the time of capture in consumer photos. Neurocomputing, 177, 529–542. https://doi.org/10.1016/j.neucom.2015.11.050

47.

van Velzen

J. H.

(2018). Students’ general knowledge of the learning process: A mixed methods study illustrating integrated data collection and data consolidation. Journal of Mixed Methods Research, 12(2), 182–203. https://doi.org/10.1177/1558689816651792

48.

Vermaak

de Klerk

H. M.

(2017). Fitting room or selling room? Millennial female consumers’ dressing room experiences. International Journal of Consumer Studies, 41(1), 11–18. https://doi.org/10.1111/ijcs.12294

49.

Vilnai-Yavetz

Tifferet

(2015). A picture is worth a thousand words: Segmenting consumers by facebook profile images. Journal of Interactive Marketing, 32, 53–69. https://doi.org/10.1016/j.intmar.2015.05.002

50.

Vogl

(2019). Integrating and consolidating data in mixed methods data analysis: Examples from focus group data with children. Journal of Mixed Methods Research, 13(4), 536–554. https://doi.org/10.1177/1558689818796364

51.

Warde

(2005). Consumption and theories of practice. Journal of Consumer Culture, 5(2), 131–153.

52.

Wills

Meah

Dickinson

A. M.

Short

(2013). Domestic Kitchen Practices: Findings from ‘kitchen life’ study University of Hertfordshire. https://doi.org/10.18745/PB.19588.

53.

Woolley

C. M.

(2009). Meeting the mixed methods challenge of integration in a sociological study of structure and agency. Journal of Mixed Methods Research, 3(1), 7–25. https://doi.org/10.1177/1558689808325774

54.

Wuggenig

(1989). Die Photobefragung als projektives Verfahren. In Hoffmann-Nowotny

H.-J.

(Ed.), Kultur und Gesellschaft: Gemeinsamer Kongress der Deutschen, der Österreichischen und der Schweizerischen Gesellschaft für Soziologie, Zürich 1988 ; Beiträge der Forschungskomitees, Sektionen und Ad-hoc-Gruppen (Vol. 16). Seismo-Verl.

55.

Kleih

A.-K.

Lehberger

Sparke

(2020). The photo-based qualitative interview ‐ potential applications to market research and current challenges. PraxisWISSEN Marketing (1), 1–11.

56.

Kleih

A.-K.

Sparke

(2021). Visual Marketing: The Importance and Consumer Recognition of Fruit Brands in Supermarket Fruit Displays. Food Quality and Preference, 104263. https://doi.org/10.1016/j.foodqual.2021.104263

57.

Zhang

Chen

(2019). Discovering the tourists’ behaviors and perceptions in a tourism destination by analyzing photos’ visual content with a computer deep learning model: The case of Beijing. Tourism Management, 75, 595–608. https://doi.org/10.1016/j.tourman.2019.07.002