Consumer’s Re-Participation in Online Flash Sale after Service Failure: A Moderated-Mediation Study

Figures & data

Figure 1. Conceptual model.

Table 2. Scale items and psychometric properties.

Scale and items	Factor loadings	CR	AVE	Cronbach’s α
Deal proneness		0.889	0.669	0.831
“Participating and taking advantage of flash sale deals makes me feel good”.	0.696
“When I participate in flash sale and take advantage of flash sale deals, I feel that I am getting a good deal”.	0.890
“Beyond the money I save, participating in flash sale and taking advantage of flash sale deals gives me a sense of joy”.	0.860
“I am more likely to buy brands that have promotional deals”.	0.813
Re-participation intention		0.863	0.667	0.762
Not likely/Likely	0.810
Improbable/Probable	0.851
Impossible/Possible	0.807
Consumer competition		0.895	0.589	0.859
“I will be competing with others to buy mobile phone during the online flash sale”.	0.676
“During the online flash sale. other potential buyers of the mobile phone are ‘rivals’ of mine”.	0.798
“I will feel like a ‘winner’ if I able to buy a mobile phone during online flash sale”.	0.866
“I will have to battle with other interested shoppers during online flash sale to buy the mobile phone”.	0.803
“I will need a ‘strategy’ or a ‘game plan’, to successfully buy mobile phone during online flash sale”.	0.700
“Trying to buy mobile phone during online flash sale feels like competition”.	0.747

Bold: Composite reliability is a measure of internal consistency in scale items. It’s also known as construct reliability. Composite reliability values were in the range of 0.863 to 0.895, all above the prescribed value of 0.70, as suggested by Nunnally (1978), indicating a good reliability of all the constructs.

Table 1. Discriminant validity.

	CR	AVE	CC	DP	PI
Consumer competition (CC)	0.895	0.589	0.768
Deal proneness (DP)	0.889	0.669	0.756*	0.818
Re-participation intention (RPI)	0.863	0.677	0.732*	0.796*	0.823

* p<.0001.

Bold: The Fronell-Larcker criterion is one of the most popular techniques used to check the discriminant validity of measurements models. According to this criterion, the square root of the average variance extracted by a construct must be greater than the correlation between the construct and any other construct. According to Fornell and Larcker (1981), the square root of the average variance extracted (AVE) should be greater than the correlation coefficient. This means that the values on the diagonal line should be greater than the values below or to the left of them in the chart.

Table 3. Summary of results.

Model		Unstandardized coefficient (beta)	Sig.	R^2	Adjusted R^2
Constant		0.100	0.751	0.459	0.450
Deal Proneness		0.685	0.000
Gender		0.093	0.407
Age		−0.01	0.385
Mediation analysis: Deal proneness → Re-participation intention via consumer competition
Serial No.	Mediating variable	Effect	SEboot	Lower limit of confidence interval LLCI	Upper limit of confidence interval LLCI
1	Consumer competition	0.2251	0.0697	0.0875	0.3626
Moderated analysis: Deal proneness → Re-participation intention (moderator deal type on main effect)
Deal proneness → Re-participation intention for		Effect	SEboot	Lower limit of confidence interval LLCI	Upper limit of confidence interval LLCI
Incidental deal type		0.6383	0.0748	0.4908	0.7858
Planned deal type		0.3148	0.0801	0.1569	0.4728
Deal proneness × deal type		−0.3235	0.1096	−0.5396	−0.1073
Moderated mediation analysis: Deal proneness → Consumer competition → Re-participation intention (moderator deal type)
Deal proneness → Consumer competition → Re-participation intention for		Effect	SEboot	Lower limit of confidence interval LLCI	Upper limit of confidence interval LLCI
Incidental deal type		0.1481	0.0495	0.0619	0.2258
Planned deal type		0.0887	0.0324	0.0340	0.1592
Index of moderated mediation
		Effect	SEboot	Lower limit of confidence interval LLCI	Upper limit of confidence interval LLCI
Deal type		−0.0594	0.0310	−0.1312	−0.0127
Moderated analysis: Deal proneness → Re-participation intention (moderator consumption motivation on main effect)
Deal proneness → Re-participation intention for		Effect	SEboot	Lower limit of confidence interval LLCI	Upper limit of confidence interval LLCI
Hedonic motivation		0.3910	0.0883	0.2168	0.5633
Utilitarian motivation		0.7321	0.0604	0.6130	0.8513
Deal proneness × Consumption motivation		0.3411	0.1070	0.1300	0.5522

Figure 2. Moderation effect of deal type and consumption motivation. Notes: (a) Moderation effect of deal type. (b) Moderation effect of deal type (on consumer competition mediation effect). (c) Moderation effect of consumption motivation.