and Wang, 2012), business simulation games (Tao et al., 2009), open source software (Leea et al., 2009), online banking (Montazemi and Qahri-Saremi, 2015), Apps stores (Rezaei et al., 2016c) and many more. Fig. 1 shows the theoretical research model.
Fig. 1. Theoretical research model. Note: Attitude towards online food delivery services (AODS).
Behavioral intention towards online food delivery services (BIOFDS).
H1. There is a positive relationship between hedonic motivation and convenience motivation.
H2. There is a positive relationship between hedonic motivation and post-usage usefulness.
H3. There is a positive relationship between prior online purchase experience and convenience motivation.
H4. There is a positive relationship between prior online purchase experience and post-usage usefulness.
H5. There is a positive relationship between time saving orientation and convenience motivation.
H6. There is a positive relationship between time saving orientation and post-usage usefulness.
H7. There is a positive relationship between price saving orientation and
convenience motivation.
H8. There is a positive relationship between price saving orientation and post-usage usefulness.
H9. There is a positive relationship between convenience motivation and post-usage usefulness.
H10. There is a positive relationship between convenience motivation and attitude towards OFD services.
H11. There is a positive relationship between convenience motivation and behavioral intention towards OFD services.
H12. Post-usage usefulness positively affects attitude towards OFD services. H13. Post-usage usefulness positively affects behavioral intention towards OFD services.
H14. There is a positive relationship between attitude towards OFD services and behavioral intention towards OFD services.
3.Research method and data analysis
To conduct empirical research and test the proposed model (Fig. 1), the positivism philosophy has been adopted along with deductive approach to theory development. Cross sectional data collection approach-survey strategy, in regards to this research, the chosen paradigm was the positivist approach and the philosophy of positivism as the collection of data is prepared to search regularities, patterns and causal relationships to create generalizations about them. According to Saunders et al. (2009), to ensure that there is no biasness in the research, the positivist approach ensures as much as possible that the data collected is neutral and not influenced by a researcher's “feelings”. Moreover, non-probability sampling method was selected. Non-prob-ability sampling refers to the selection of respondents being a subjective judgment by the researcher (Saunders et al., 2009; Sekaran and Reviews, 2006). This approach is selected because the population is not known, the sampling frame is not available, and data collection can be done more quickly at a lesser cost in marketing research (Kotler and Armstrong, 2015). In addition, convenience sampling technique
is appropriate to collect samples as it allows the ease of data collection as well as at a relatively low cost (Battaglia, 2008). According to Saunders et al. (2009), although selected haphazardly, convenience sampling is able to meet the purpose of the research objectives. To avoid problems that may arise from convenience sampling, a larger variation in the sample was selected. Thus, the purposive sampling technique was applied in this research. Thus, purposive sampling or more commonly known as judgmental sampling is selected to collect the required data. Purposive sampling is used because it selects samples that are more representative of the population. Thus, it has been ensured that all participants selected know what attitude towards OFD services is and are users of attitude towards OFD services to give a better and more accurate representation of the population, purposive sampling was used.
Questionnaires are used in the data collection process for this research. Questionnaires were selected due to various advantages such as cost advantage, greater geographical coverage, provides anonymity and also to reduce the biasness through the pressure an interviewer can give. A pre-test with 30 samples was conducted to ensure the questionnaire is suitable and usable for this research. According to previous study (Burmeister and Aitken, 2012), a reasonable number of sample for purposive depends on the technique. In this research, the reasonable number or minimum requirement of sample size was detected 100–150 samples (Rezaei, 2015) to test the model using PLS-SEM. However, other research has found the effect of larger sample sizes to give higher accuracy (Delice, 2010). The questionnaire was designed in two sections. In the demographics section, there were a total of seven variables: gender, age, nationality, ethnicity, highest level of education, attitude towards OFD services usage frequency, and Usage Duration. The second section of questionnaire include the research variables that were adopted from previous studies. The scales used in second section was nominal, as the 5 point Likert scale was primarily adopted. It ranges from Strongly Disagree to Strongly Agree (5). According to Dawes (2007), a 5 or 7 point Likert scale is more likely to produce slightly higher mean scores as compared to a 10 point scale, which makes comparing data a much easier process.
For the main data collection, the hybrid distribution method was considered. Firstly, paper-pencil questionnaire was distributed to potential respondents. Based on the sampling method, potential respondents were selected and to complete the questionnaire. Participation were strictly voluntary and anonymous. Secondly, online questionnaire with paper-pencil questionnaire was distributed via hyperlink in e-mails and on social networking sites. Respondents were given one week to complete the questionnaire and responses captured instantly upon the completion of the questionnaire. A gentle reminder was given to those who have yet to complete the survey after three days of the first request. 100 hardcopies were distributed and collected while 200 invitations were sent out on social media site Facebook to collect response, as only 50% were expected to be collected. The samples collected were from students. Thus, hardcopies were distributed in the vicinity of the campus, while online questionnaires were sent to students of different universities off campus. All questionnaires were distributed in a three-day span to ensure consistency in the sampling time frame. Furthermore, as OFD services were limited within the Klang Valley, the respondents were also confined to this same area. Of the 236 responses collected, there were 12 that were incomplete and thus deemed unusable. A total of 95% of the collected response is usable.
4.Discussion and conclusion
The TAM, model of IT Continuation and Contingency Framework were used to investigate the relationship between external factors and intention towards OFD services mediated by post-usage usefulness and convenience motivation. Davis (1989) hypothesized that perceived usefulness and ease of use will determine attitudes towards a technology. Results have further demonstrated the validity of the theoretical foundations used. It has also shown the relationship between latent variables towards attitude and behavioral intentions. This means that with a better perception of post-usage usefulness and convenience motivation, a person's attitude towards OFD services will improve significantly, thereby increasing intentions to use OFD services. Therefore, H1 and H2 and H3 are accepted.
相关推荐：
loading