Your question comes up frequently in the field of innovation.  By and large, multidimensional innovation is preferred unless your research deals with a specific and very narrow focused innovation. For example, in the article 

García-Zamora, E., González-Benito, O., & Muñoz-Gallego, P. A. (2013). Organizational and environmental factors as moderators of the relationship between multidimensional innovation and performance. Innovation, 15(2), 224-244., the authors claim that:

"Researchers have explored various perspectives to understand the nature of innovation. Today, multidimensional approaches challenge one-dimensional methods, with the proposition that each type of innovation exhibits its own characteristics. Typologies of innovation have been proposed since the 1970s (Zaltman, Duncan, & Holbek, 1973)." and "Maravelakis, Bilalis, Antoniadis, Jones, and Moustakis (2006) measure organizational innovation according to three areas: product, process, and administration. Lin and Chen (2007) add innovation and strategic marketing to the list."

Certainly individual level innovative work behaviour can be measured using multidimensional scaling unlike, many qualitative reforms are attempt measurement using multidimensional scaling.

the variables received from literature review and in-depth  interviews with experts can be executed on samples to check reliability of variables as a pilot testing. Finalized set of variables executed on samples need to be analyzed using exploratory factor analysis. factor analysis would provide number of categories of independent variables which determines innovative work behaviour.

In line with Bhola I refer exploratory factor analysis (EFA) to determine individual-level innovative work behavior (IWB) by developing scale based on multi dimensional scale. Finally, go for criterion validity. 

Michael,

In your review of the individual innovative work behavior construct you should take a look at the work of de Jong and den Hartog.  IWB sub-dimensions include opportunity exploration, idea generation, idea championing and idea implementation.  The first two sub-dimensions are typically the more creative dimensions and may be termed the ideation dimensions.  The latter two sub-dimensions convert creative ideas into useful realizable benefits ie into innovations.  It is therefore difficult to capture all these dimensions in a unitary construct.  There is still a measurement challenge with strong valid 4- factor scales but certainly valid 2-factor scales do exist.

Refer;

https://www.researchgate.net/profile/Deanne_Den_Hartog/publication/228254317_Measuring_Innovative_Work_Behaviour/links/09e4150e83a0222c40000000.pdf

http://www.entrepreneurship-sme.eu/pdf-ez/h200820.pdf

 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1555939

Silburn

Article Measuring Innovative Work Behavior

Hi Michael,

It really depends on your research objectives.  For example, if you are measuring individual innovativeness as the outcome of environmental factors, it may be enough to use a uni-d scale.  But, on the other hand, if innovative behavior is treated in your study as a factor which affects organizational outputs such as sustainable development, then it will be better to use a multi-d scale to measure it.  That will allow you to identify later which particular behavior has a stronger effect on the output.

Hope that helps.

Aida

Thanks everyone for the advice. The specific challenge i am facing is that in one study Jansens scale emerges as multidimensional after an EFA but with moderate to high levels of intercorrelations between the seperate dimensions. In another sample i couldnt get a multi-dimensional structure as the EFA showed a single factor structure. Anyones thoughts on this

Have you considered the construct, innovation process performance? It is an individual based construct.

See the following article:

Rampersad, G., Plewa, C., & Troshani, I. (2012). Investigating the use of information technology in managing innovation: A case study from a university technology transfer office. Journal of Engineering and Technology Management, 29(1), 3-21.

http://www.sciencedirect.com/science/article/pii/S0923474811000403

Hope it helps.

Cheers,

Giselle

I agree with Sandra's answer: multidimensional measures of innovation is preferred unless your research deals with a specific and very narrow focused innovation.

With some exceptions (Ebersberger et al. 2011), most analyses on innovation analyse a single R&D variable, usually internal R&D expenditures. However, certain empirical studies on this question suggest the need to approximate innovation from a variety of angles (AnniqueUn and Romero –Martínez 2009; Vega-Jurado et al. 2009). Some firms may concentrate their technological effort on aspects other than developing internal R&D capabilities (Franco and Quadros 2003; Schmidt and Sofka 2009; Cantwell, 2014; Narula, 2002). This suggests that the approach to R&D needs to be comprehensive, perhaps taking into consideration all innovative activities in terms of Oslo Manual.

On the other hand, there are significant differences across industries (NACE rev) in innovative activities that are developped and in their relative importance (Garcia, Molero and Rama 2014, 2015). So, I suggest to construct some measure referred to industry average in order to avoid industry bias.