Considerations in apply the Semantic Differential Technique

Semantic differential (SD) has been widely used to measure attitudes since Osgood et al. (1957) proposed this method in the 1950’s. Its broad application is attributable to its relative low entry barrier in terms of requisite training and costs. However, like most of the other methods, there are considerations and issues for using it properly and for deploying it where it is useful. This post is to discuss some of these considerations and issues.

One of the most important things to consider is culture, especially when SD involves emotive evaluations. The empirical research Scherer and Wallbott (1994) conduct with data from 37 countries does not focus on SD per se, but it finds a “high degree of universality of differential emotion patterning and important cultural differences in emotion elicitation, regulation, symbolic representation, and social sharing.” (326) It suggests that culture plays a certain role in some dimensions regarding emotions. However, they are also careful in postulating whether some emotions might be unique to a certain culture. On the other hand, Al-Hindawe (1996) cites a study by Furuya-Nakajima and Vogt (1990), which finds significant cultural difference in how “ambition” and “self-confidence” are considered differently in Japan compared to Western cultures. Al-Hindawe (1996) also indicates gender as a potential variable influencing how traits are evaluated.

Additionally, Bradley and Lang (1994) caution that relying on a verbal rating system makes the technique difficult to use in non-English speaking countries and in populations not linguistically sophisticated (e.g., children or people with linguistic disorders.) There are other studies in the literature that concur this critique (e.g., Heise 1970). Bradley and Lang (ibid.) study a picture-oriented instrument called the Self-Assessment Manikin (SAM) and find it easier to administer than the traditional SD and overcoming some of the said pitfalls.

Are emotions suitable traits to use in SD? There are several sub-issues under the “emotion” umbrella. For example, how do we define “emotion?” Morgan & Heise (1988: 19-20) distinguish pure emotion (“internal, mental feeling whose focus is solely on affect”) from traits (e.g., trustworthy, warmhearted), physical states (e.g., sleepy, droopy), and cognitive states (e.g., alert, confused). However, in some cultures and contexts, some of these traits, physical states, or cognitive states may imply some emotional state. For example, sleepiness might indicate emotional detachment. In other words cultural and emotional considerations can be intertwined.

Another sub-issue related to emotion is when to use emotional traits in SD. I find two lines of research with bearing on this topic: Vergara et al. (2007) argue that commercial products are “bearers of users’ feelings” (9), so tools as mundane as hammers can be considered “emotional designs” as well and be potential candidates for studying with the SD. The other relevant study I find is by Allen et al. (1992), who find emotive reports and attitudinal judgments can be complementary in their study on behavioral predictors. For example, they think products that evoke nostalgic experience are good candidates for benefiting from taking emotions into account.

Other than cultural considerations and applicability of emotional traits, there are other methodological issues in using SD as well. For example, Bradley and Lang (ibid.) critique that using the SD technique involves analyzing a large database that requires statistical expertise such as factor analysis. Moreover, how do we know what people mean by choosing zero on the scale? It seems to me it could mean indifference, not applicable, don’t know, and refuse to answer (for whatever reason). Furthermore, we also need to consider potential threats to validity of what we measure with the SD instrument. For example, Heise (ibid.) suggest social desirability might plague participants’ responses.

In conclusion, SD is applicable in many contexts and situations, but we need to be aware of the potential issues that might compromise its usefulness. To deploy the method wisely, we also need to keep up with the methodological development of this method becomes more sophisticated.

References

1. Al-Hindawe, J. 1996. Considerations when constructing a Semantic Differential Scale. Bundoora Victoria, AU: LaTrobe University.
2. Allen, C.T., K.A. Machleit, and S.S. Kleine. 1992. A Comparison of Attitudes and Emotions as Predictors of Behavior at Diverse Levels of Behavioral Experience. Journal of Consumer Research 18: 493-504.
3. Bradley, M.M. and P.J. Lang. 1994. Measuring Emotion: The Self-Assessment Manikin and the Semantic Differential. Journal of Behavioral Therapy & Experimental Psychiatrics 25(1): 49-59.
4. Heise, D.R. 1970. The Semantic Differential and Attitude Research. In F. Summers (Ed.) Attitude Measurement, Chapter 14, pp. 235-253. Chicago, IL: Rand McNally.
5. Morgan, R.L. and D.R. Heise. 1988. Structure of Emotions. Social Psychology Quarterly 51(1): 19-31.
6. Osgood, C. E., P. H. Tannenbaum, and G. J. Suci. 1957. The Measurement of Meaning. Urbana: University of Illinois Press.
7. Scherer, K.R. and H.G. Wallbott. 1994. Evidence for Universality and Cultural Variation of Differential Emotion Response Patterning. Journal of Personality and Social Psychology 66(2): 310-328.
8. Vergara, M., S. Mondragón, J.L. Sancho-Bru, P. Company, and A. Pérez-González. 2007. User Profile Differences in Semantic Design - Application To Hand Tools. Presented at the International Conference On Engineering Design, Iced’07, 28 - 31 August 2007, Cite Des Sciences Et De L'industrie, Paris, France.

Artifacts and Meanings

Akama et al. suggest using artifacts in interviews as “triggers for reflection and imagination, tools for the articulation and communication of ideas and experience, and facilitators for participation and generative meaning-making” (2007: 173). Their discussion about artifacts reminds me of the research Csikszentmihalyi (1993) does on the meanings of household objects. Considering the rich underlying meanings and social relationships, artifacts could be a powerful tool to facilitate interaction, retrieve memories, and encourage insights when coupled with traditional interview techniques.

Artifacts provide “a form of comfort” (Akama et al. 2007: 177) especially when they are indigenous to the participant’s native environment, such as their home. The participant’s familiarity with the artifacts makes the interview process feel more relaxed and less formal. Therefore, artifacts can be good interview ice-breakers that invite participation and bridge the psychological gap between the participant and the interviewer.

Artifacts also provide focal points in the interview process. The material dimension allows for sensory touch, which can aid participants in processing thoughts and triggering memories of experience with the artifacts. They enable recalling of histories and stories associated with them, as well as other objects they have relations with. There can also be other people involved in these histories, stories, memories, and network of objects. With proper follow-up questions, artifacts serve as starting points that can help the interviewer snowball questions, thereby catalyze generative meaning-making, nuances and insights that otherwise may not be uncovered.

Focusing attention on artifacts enables opportunities for interview participants to reexamine how they relate to the objects at issue. In so doing, they reveal their sense about who they are. As Csikszentmihalyi (1993) argue,

[Objects help] both focus attention, reducing entropy in consciousness, and vividly brings back old memories and experiences, thus adding a sense of depth and wholeness to the self of its owner… the most meaningful symbol of his private self … had the power to put him back in touch with himself. (25)

Therefore, artifacts “embody the values and tastes as well as the accomplishments of the owner” and they serve as “repositories of meanings about the self” (Csikszentmihalyi 1993: 25-26).

The way in which people talk about artifacts they own also reveal how they place themselves in the social networks in which they define themselves. For example, the participant I interviewed today selected a ceramic bean pot and a cookbook (see photo above) from his grandmother. The cookbook was published in 1903 and is in Norwegian. It reveals the participant’s family origin, and his relationship with his grandmother, who laboriously prepared very time-consuming and yet delicious Christmas meals that the participant still vividly remembers more than half a century later. 

As repository of meanings, artifacts are often capable of drawing out rich narratives complete with sounds, smell, images, tastes, feelings, emotions, and the context in which these sensory dimensions exist. Therefore, engaging interview participants with artifacts opens the window to peek into the holistic experience the owner of the artifacts has had.

To me, the most important take-away of studying the use of artifacts in interviews is developing sensitivity toward the interaction between artifacts and people, and toward the experience surrounding such interaction. Artifacts are not just objects. The acts of acquiring, keeping, and using an artifact, as well as giving it a spot in their private homes all have meanings. Why and how an artifact becomes a part of someone’s life is intriguing.

References:
Akama, Y., R. Cooper, L. Vaughan, and S. Viller. 2007. Show and Tell: Accessing and Communicating Implicit Knowledge Through Artefacts. Artifact I (3):172–181.

Csikszentmihalyi, M. 1993. Why We Need Things. In S. Lubar and W.D. Kingery (eds.) History from Things: Essays on Material Culture , pp. 20–29. London: Smithsonian institution press.