Maybe add an indirect measure that is well-suited to capture emotional changes in daily life (see references below)?

Lischetzke, T., Cuccodoro, G., Gauger, A., Todeschini, L., & Eid, M. (2005). Measuring Affective Clarity Indirectly: Individual Differences in Response Latencies of State Affect Ratings. Emotion, 5(4), 431-445. doi: 10.1037/1528-3542.5.4.431

Lischetzke, T., Angelova, R., & Eid, M. (2011). Validating an indirect measure of clarity of feelings: Evidence from laboratory and naturalistic settings. Psychological Assessment, 23(2), 447-455. doi: 10.1037/a0022211

Best,

Ramzi

Many others have suggested incorporating RTs. RT measures in Lischetzke et al (2011) are proposed to be ab indicator of an individual's clarity of feelings, not necessarily a good way to measure emotional fluctuations. 

Plus, I think Rita's question was more about methods (other than MSSD) of capturing emotion fluctuations given the self-report data  she already has.

There's a relatively new app called the Moodometer, which is an "interactive mood diary for monitoring and understanding emotional wellbeing". It's along the lines of a diary study, but can be run on multifunction devices like phones, and may be better equipped for identifying relationships between mood fluctuations and a range of causes.

http://myhealthapps.net/app/details/363/moodometer

You can try to use the Emotional Processing Scale by Roger Baker, available from Hogrefe Publishing.

Perhaprs they can evaluate they feelings in valence and arousal right after they report the event using the Self Assessment Manikin.

Bradley, M. M., & Lang, P. J. (1994). Measuring emotion: the self-assessment manikin and the semantic differential. Journal of behavior therapy and experimental psychiatry, 25(1), 49-59.

Morris, J. D. (1995). Observations: SAM: the Self-Assessment Manikin; an efficient cross-cultural measurement of emotional response. Journal of advertising research, 35(6), 63-68.

Hi all Thanks for your suggestions.

I already collected data, and have information on the valence and arousal (I used the SPANE scale). What I was asking was for methods of analyzis to ascertain the influence of emotional fluctuation on the outcomes.

MSSD is considered the state of the art procedure for calculating affect variability (lability) assessed using daily diary, EMA, or ESM data. Thus, I would suggest using that. However, others have used a personalized SD approach that produces similar but different estimates from event level data. A good description is located in the attached article by Gottfredson & Hussong (2013).

Because Rita asked for creative analysis method, I was going to suggest using techniques employed in timeseries analysis (i.e. brain activation over time).

If you can reveal what your hypotheses are, perhaps we can have a more informed discussion.

But without know that,  what I'm concerned with is that in this case, sampling was done once daily, and it was not random or time-locked, but instead it was linked to specific events that were different in nature for everybody, and presumably occurred at different times of the day.

So, the sampling rate and sampling trigger may be a problem, depending on your hypotheses, and whether these the hypotheses involve individual or intra-individual effects. 

Rita. I have a couple of suggestions:

a) about collecting your data. Diaries could be a right choice when you have high motivated participants, with self-observation skills (interoception) and willing to perform them (high adherence), on the other side you may try the EAR method (Mehl, et al., DOI: 10.1177/0956797610362675) you will collect a wide set of socially relevant data, track frequency, intensity and maybe the exact point of emotional change along the course of daily living .

b) once you have data sets . Statistical analysis coming from dynamical assessments have used ANOVA (Burden, 2010) in order to analyze long lasting data from groups in spite that this kind of method no always satisfy the basic assumptions (Jennings & Gianaros, 2007). Further this kind of analysis is not sensitive for discrete changes or time limited events episodes that become expressed along a non identifiable time embedded within a longitudinal psycho physiological data sets (Rosenfield, Zhou, Wilhelm, Conrad, Roth & Meuret, 2010). In fact statistical inference test are not designed to identify the specific moment when change emerge within a time series (Chen & Gupta, 2012). Then you need an statistical analysis option for time series that let you detect the specific time of change with its magnitude and duration. One option is the Point of Change Analysis, an adaptation coming from statistical control quality theory ; it assumes that any production process include natural variability or background noise , coming from the accumulated effects of small causes; if a process work only with this kind of variations can be called stable (under statistical control) , when a process include several variation causes outside the of the stable process , then it is labeled an unstable one , therefore the main objective of statistical control process is fast detection of variations in order to label them as stable or un stable (Woodall, 2000; Woodall & Montgomery, 1999) .

I have recently finished developing a set of nonverbal mood scales Dynamic Visual Analog Mood Scales (D-VAMS). They comprise a brief (< 5 minutes), self-report measure covering the principle mood types, and are based on images of facial expressions which can be modulated by the use of sliders (see dvams.com). They are primarily designed with people with communication difficulties in mind, but their language independence makes them useful if you are using people who do not speak English. The instrument is currently undergoing validation, but the scales are based on normative judgement data from a series of 5 experiments conducted over the last 3 years (see xvams.com for more details). There is a poster on my page that explains their design.

Gottschalk (Lewis A.) content analysis of verbal behavior. (Now available as software called Picard 2000