TECHNOLOGY ACCEPTANCE MODEL (TAM) FACTORS AND SOCIAL FACTORS ANALYSIS THROUGH ATTITUDE TOWARDS TO USE ON INTENTION TO PURCHASE OF KISAH KITA NGOPI ONLINE CAFÉ

The covid-19 pandemic has forced entrepreneurs to switch to online sales, including Kisah Kita Ngopi (KKN) Café. It is essential to know the factors that affect a customer’s Intention to Purchase when online, such as the TAM Factors (perceived usefulness, perceived ease of use) and social factors (social influence, peer influence) attitude towards to use in using online applications. This research uses the SEM data analysis method assisted by Partial Least Square (PLS) software. Data collection is carried out by distributing questionnaires in the form of Google forms to people in Sidoarjo and Surabaya who have made online purchases at the KKN Café. In this study, the random sampling technique obtained 85 respondents. The results of this study indicate that (1) perceived usefulness affects attitude toward mobile app use, (2) perceived ease of use affects attitude toward mobile app use, (3) social influence does not affect attitude toward mobile app use, (4) peer influence affects attitude towards mobile app use, (5) attitude towards mobile app user does not affect intention to purchase, (6) perceived usefulness does not affect intention to purchase, (7) perceived ease of use does not affect intention to purchase, (8) social influence does not affect intention to purchase, and (9) peer influence does not affect intention to purchase. Therefore, based on the results of this study, it can be concluded that the TAM factors scores are higher than those related to social factors with regards to attitude towards to use.


INTRODUCTION
In January 2020, of the total 7.75 billion population, 4.54 billion used the internet and 5.19 billion used smartphones. People use the internet because the *Corresponding Author. e-mail: askalis.dioa@ieu.ac.id idea of the internet itself is to make things easier, and it has become a basic human need. Of the more than 200 countries in the world, in Iceland, Kuwait, Qatar, United Arab Emirates, and Bahrain, the penetration rate of internet usage reaches 99%, which means that almost all citizens can easily access the internet (Ramadhan, 2020). During the COVID-19 pandemic, including in Indonesia, there was a change in internet usage behavior. The Indonesian Internet Service Providers Association (APJII) surveyed the penetration and behavior of internet users in Indonesia from 2019 to the second quarter of 2020. This survey shows that there is a change in the pattern of internet users' usage in Indonesia, namely the devices used, whereby 95.4% of the internet users used a smartphone to connect to the internet. The APJII survey results also show that most respondents use smartphones for eight hours. Internet users in Indonesia, especially Java Island with 56.4%, and Sulawesi island with 7% (Damar, 2020). The technology acceptance model (TAM) is implemented to predict customer interest and greater validity in adapting to new technologies. TAM is one of the most successful models in examining Technology acceptance and adaptation in customer behavior regarding technology. TAM also provides a perception of acceptance of technology, which in the context of this research is the use of smartphones. In the use of smartphones, benefits (perceived usefulness) and ease (perceived ease of use) in using smartphones are taken as the factors for user perception and behavior and as a benchmark in smartphone acceptance. The usefulness and convenience perceived by users can contribute to attitudes and behavioral intentions toward new technology. Research results from (Vahdat et al., 2020) showed that PU does not significantly impact attitudes toward mobile apps. However, PEOU, SI, PI, and Intention to purchase positively affect attitude. Moderation analysis also shows that the age factor is the factor that has mediating effect on the impact of PU towards attitude towards use.
Kisah Kita Ngopi Cafe is a cafe located on Jalan DPR III Pagewojo Sidoarjo, founded in early 2019 and operating in September 2019. The concept offered in Kisah Kita Ngopi cafe is to build social identity and simultaneously encourage joint activities with various communities (special factor). Café KKN must also do business changes to anticipate the change in customers' purchasing behavior during the Covid19 pandemic, which is switching to online purchases, by providing online sales through various media. Thus, the formulation of the problems proposed based on the discussion that has been carried out above are (1) Does perceived usefulness affect intention to purchase in KKN Café online sales? (2) Does perceive ease of use affect intention to purchase at KKN Café online sales? (3) Does social influence affect intention to purchase in KKN Café online sales? (4) Does peer influence affect intention to purchase in KKN Café online sales? (5) Does attitude affect intention to purchase in KKN Café online sales? (6) Does perceive usefulness affect intention to purchase with attitude as a moderator in KKN Café online sales? (7) Does perceive ease of use affect intention to purchase with attitude as a mediator in KKN Café online sales? (8) Does social influence affect intention to purchase with attitude as a mediator in KKN Café online sales? (9) Does peer influence affect intention to purchase with attitude mediator in KKN Café online sales?

LITERATURE REVIEW
This research is based on theory planned behaviour (TPB) (Ajzen, 1991) which discusses the limit of the actual model in handling behavior whereby people have incomplete control and desire. The majority of the behavior is not entirely under control; behavioral control results from the control belief that underlies and relates to resources, opportunity, experience, and information about others' experiences. Attitude, subjective norm, and behavioral control, especially toward the behavioural intention.
This research is based on the Technology acceptance model (TAM) was researched by Davis (1989), adapted from the theory of reasoned action (TRA) (Fishbein & Ajzen, 1975) in psychology research. The TRA argues that a behavioral goal drives character behavior, wherein the behavioral aim is a function of a person's mindset towards the behavior and subjective norms surrounding the performance of the behavior. In other words, it states that one behavior and the intent to behave is a feature of 1's three mindset towards the behavior and their perceptions about the behavior. Therefore, conduct is the feature of each attitude and ideal. Meanwhile, TAM proposes that perceived ease of use and perceived usefulness of technology are predictors of a person's mindset toward using the technology, subsequent behavioral intentions, and actual usage. Perceived ease of use changed into also considered to persuade perceived usefulness of generation.
Intention to purchase is the measure of an individual's intention to act with certain behavior in deciding to purchase a product or service. Intention to purchase is an individual's willingness to purchase a good or service (Doan, 2020). To measure intention to purchase variable, three indicators are needed (Putrevu & Lord, 1994), which are: (1) interest to acquire the product, (2) considering making a purchase, (3) interest to try.
Individuals' intentions to carry out activities in the context of mobile apps are determined by their attitudes towards applications in general (Carter & Yeo, 2016). To measure attitude towards mobile apps variable, five indicators are needed (Chen Ying et al., 2015), namely: (1) the attitude of acceptance of technology, (2) feeling about the benefits of technology, (3) feeling the value of a technology, (4) a love for technology, (5) a pleasant experience using technology.
Perceived ease of use refers to the extent to which people believe that technology is easy to use (Veríssimo, 2016) and can also be seen as a relative advantage perceived by the user so that it is considered a construction or a relative advantage that can be said as better than its predecessor (Muñoz-Leiva et al., 2017). Perceived ease of use variable can be measured using four indicators (Kucukusta et al., 2015), which are: (1) easy to learn, (2) less requirement of mental effort, (3) being simple, (4) easy to follow instructions.
Social influence is the measure of how strong human social interaction is. Social influences can arise either when individuals are being persuaded to get involved because of pressure from friends and family or when they want to impress others (Shaw & Sergueeva, 2019). Social influence variable can be measured using six indicators (Wang & Lin, 2011), which are: (1) substantial influence on the seller, (2) substantial influence on the buyer, (3) social influence effect, (4) people with the same interest influence, (5) someone important influence, (6) positive social feedback.

Methods
Figure 2 explains the analysis model in this study with the first independent variable, namely TAM factors, consisting of perceived usefulness (X1) and perceived ease of use (X2). The second independent variable is Social factors, consisting of Social influence (X3) and Peer influence (X4). The two independent variables affect the dependent variable. The first dependent variable is attitude towards use (Y1), which is a mediator in this study, and the second dependent variable is Intention to purchase (Y2). The population in this study is people who live in Surabaya and Sidoarjo who have shopped online at KKN Café. The sample is part of the total of population with specific characteristics consisting of several selected members. A sampling technique is used for sample selection to determine the research sample (Triyono, 2018). The sampling technique used in this study is non-probability sampling by purposive sampling method. The samples in this research are people who have lived in Surabaya and Sidoarjo. These people have shopped online through GoFood, Grab, Shopee, and aged 15 and above. The number of samples was set as many as 85 respondents. The number of samples is based on an unknown population or infinite population so Davis & Coenza (1993) formula is used as follows: This study used a primary data measurement scale, an interval scale. There are seven interval scale indicators; responses with the scale of 1 indicate strongly disagree to 7 indicate strongly agree. This study used Partial Least Square (PLS) quantitative analysis approach. Partial Least Square (PLS) becomes a good alternative when the sample size is small. Validity and reliability tests were conducted to determine the accuracy and reliability of the questionnaire. The data processing using SmartPLS (Partial Least Square) 3.0 software.

Validity Test
Validity test shows the degree of acceptance of research results by the community with certain criteria. In this study, internal validity and external validity is used. Two types of validity assessment commonly applied in PLS-based data analysis are convergent validity and discriminant validity.

Reliability Test
Reliability test ensures the extent of error or no bias and ensures the measurement is consistent for various items of the instrument. The reliability test in this study, which examined the consistency of the research instrument, used Cronbach's alpha values.

Structural Model (Inner Model)
Structural model conducts evaluation using R 2 in for dependent construct, for the significance test between constructs in structural model, there are coefficient path values or t-values for each path (Abdillah & Jogiyanto, 2015).
To measure the level of variation in the change in independent variable on the dependent variable, R 2 value is used. The higher the R 2 means the better the prediction of the proposed research.
The research model will be better or more fitting with the data if the Q Square is higher. Q Square can be measured by the formula Q Square = 1-[(1 -R1) x (1 -R2)]. If the result of Q Square > 0 it can be concluded that the model shows predictive relevance, whereas if Q Square < 0 it can be concluded that the model does not show predictive relevance (Joseph F Hair, 2014).

Hypothesis Testing
Hypothesis testing is conducted by comparing the path coefficient and outer loading between T Statistics and T-table value using the significance level of 5% or 1.96 (Hadikusuma & Jaolis, 2019). If the T Statistics is <T-tabel then it will be deemed insignificant, but if the T Statistics value >T-tabel it will be deemed significant.  Based on Table 3, the result of data processing using Partial Least Square (PLS) shows that the outer loading value > 0.7, except the indicator item in the peer influence variable, which is X4.4 with outer loading value < 0.7 of 0.68. Indicator variables with outer loading >0.7 indicate high level of validity and vice versa, therefore they fulfil the convergent validity. Thus, peer influence indicator variable, which is X4.4, need to be eliminated or removed from the model.

Final Convergent Validity
When conducting the convergent validity measurement, one thing that needs to be paid attention to is the value of each outer loading. The indicator can be deemed as fulfilling the convergent validity when the outer loading value > 0.7 and the Average Variance Extracted (AVE) value > 0.5 (Sarstedt & Christian M. Ringle, 2017) and what was obtained from the outer model evaluation result:  Table 3 shows that the outer loading values of each indicator variables above is > 0.7. Average Variance Extracted (AVE) of each variable also show values of > 0.5. Based on the data above, it can be concluded that the indicators used in this research have fulfilled the convergent validity and can conceptually measure the variable being observed.

Discriminant Validity
When conducting discriminant validity measurement, one thing that needs to be paid attention to is the cross loading value. Indicator used can be deemed to fulfil discriminant validity if it has the highest value in each indicator in the variable. Below is the table of cross loading values in the research:   Based on Table 6, the combach's alpha values of each variables is > 0.6, whereby combach's alpha of 0.6 is generally accepted, thus the reliability values can be accepted (Ursachi et al., 2015). In this research, each variables show composite reliability values of > 0.7, therefore it can be inferred that each variables in this research is reliable.

R Square (R 2 ) Value
When conducting the structural (inner) model evaluation, the first stage that is conducted is to determine the R Square (R 2 ) value. Below is the R Square (R 2 ) value obtained through data processing using SmartPLS (Partial Least Square) 3.0 software: The result of R Square (R 2 ) value on attitude towards to use is 0.781, which means that perceived usefulness, perceived ease of use, Social influence, Peer influence have influence of 78.1% on online purchases in KKN Café, whereas the variables not explained in this study account for 21.9% of impact on attitude towards to use. R Square (R 2 ) value on intention to purchase is 0.522, which imply that perceived usefulness, perceived ease of use, social influence, Peer influence, and attitude towards to use have impact of 52.2% on KKN Café online purchase, whereas the variables not explained in this study account for 47.8% of impact on intention to purchase.

Q Square Value(Q 2 )
Q Square can asses the accuracy of a prediction. If the Q Square (Q 2 ) calculation > 0, it can be concluded that the model shows predictive relevance, but if the Q Square calculation < 0, it can be inferred that the model does not show predictive relevance (Joseph F Hair, 2014). Below is the Q Square (Q 2 ) evaluation result: Based on the calculation above, the Q Square result obtained is 0.896, it can be inferred that with this value, it can predict 89.6% of attitude towards to use and intention to purchase for online purchases in KKN Café.

Result
Hypothesis testing is conducted using structural (inner) model result using SmartPLS (Partial Least Square) 3.0 software. If the t-statistics > 1.96, then the research hypothesis can be accepted with the significance level p-value < 0.05 or (5%) (Hadikusuma & Jaolis, 2019). Below is the result of hypothesis testing in this research:

DISCUSSION
The coefficient value of the effect of perceived usefulness on attitude towards to use is 0.595 and it can be interpreted as perceived usefulness having positive relationship with attitude towards to use. The t-statistics value it is equal to 5.805, where it can be concluded that the value of t-statistics> 1.96. Based on these results, it can be concluded that the first hypothesis is accepted; that is, perceived usefulness has a significant effect on attitude towards to use in online shopping at KKN Café in Sidoarjo and Surabaya. This is supported by research conducted by (Li et al., 2017) that stated that perceived usefulness shows positive effect on attitude towards to use. H1: Perceived usefulness has impact on attitude towards to use in online market place.
The coefficient value of the impact of perceived ease of use on attitude towards to use is 0.260 and it can be interpreted as perceived usefulness having positive relationship with attitude towards to use. The t-statistics value it is equal to 2.145, where it can be concluded that the value of t-statistics > 1.96. Based on this result, it can be concluded that the second hypothesis is accepted, since perceived ease of use significantly impact attitude towards to use in shopping online in KKN Café in Sidoarjo and Surabaya. Research conducted by (Muñoz-Leiva et al., 2017) also stated that perceived ease of use has positive impact on attitude of acceptance for mobile banking users. (Vahdat et al., 2020) also supported the same statement in their study. H2: Perceived ease of use has impact on attitude towards to use in online market place.
The coefficient value of the impact of social influence on attitude towards to use is -0.104 and it can be interpreted as social influence having negative relationship with attitude towards to use. Meanwhile, the t-statistics value is 0.926, so it can be concluded that the t-statistics value is < 1.96. Based on these results, it can be concluded that the third hypothesis is rejected because social influence has no significant effect on attitude towards to use in online shopping at KKN Café in Sidoarjo and Surabaya. Research conducted by (Hadikusuma & Jaolis, 2019) also stated the same; that is, social influence has no impact on attitude towards using in the context of said research, which is mobile payment. H3: Social influence has impact on attitude towards to use in online market place.
The coefficient value of the impact of peer influence on attitude towards to use is 0.267, thus it can be interpreted as peer influence having positive relationship with attitude towards to use. The t-statistics value is 3.176, therefore it can be concluded that the t-statistics value > 1.96 and thus it can be concluded that the fourth hypothesis is accepted, since peer influence has significant impact on attitude towards to use in shopping using online application in KKN Café in Sidoarjo and Surabaya. H4: Peer influence has impact on attitude towards to use in online market place.
The coefficient value of the impact of attitude towards to use on intention to purchase is 0.311 thus it can be interpreted as attitude towards to use having positive relationship with intention to purchase. Whereas from the t-statistics value of 1.775, it can be concluded that t-statistics < 1.96. Based on this result, it can be concluded that the fifth hypothesis is rejected, since attitude towards to use has no significant impact on Intention to Purchase use in online shopping at KKN Café in Sidoarjo and Surabaya. Research conducted by (Vahdat et al., 2020) stated that peer influence has no impact on attitude in using shopping application. H5: Attitude towards to has impact on intention to purchase in online market place.
The coefficient value of the impact of perceived usefulness on intention to purchase is 0.295 thus it can be interpreted as perceived usefulness not having positive relationship with intention to purchase. Whereas from the t-statistics value of 1.867, it can be concluded that t-statistics < 1.96. Based on this result, it can be concluded that the sixth hypothesis is rejected, since perceived usefulness has no significant impact on intention to purchase use in online shopping at KKN Café in Sidoarjo and Surabaya. This is supported by (Hsu & Lin, 2016) that stated that there is no significant relationship between attitude towards mobile app use with intention to purchase. H6: Perceived usefulness has impact on intention to purchase with mediation of attitude towards to use in online market place. The coefficient value of the impact of perceived ease of use on intention to purchase is -0. Thus it can be interpreted as perceived ease of use not having positive relationship with intention to purchase. Whereas from the t-statistics value of 0.677, it can be concluded that t-statistics < 1.96. Based on this result, it can be concluded that the seventh hypothesis is rejected since perceived ease of use has no significant impact on intention to purchase in online shopping at KKN Café in Sidoarjo and Surabaya. This is supported by (Muñoz-Leiva et al., 2017) that stated that perceived usefulness has no impact on intention to use. Research by (Setiawan & Sugiharto, 2014) also stated that perceived usefulness has no impact on intention to use. H7: Perceived ease of use has impact on intention to purchase with mediation of attitude towards to use in online market place.
The coefficient value of the impact of social influence on intention to purchase is 0.292. Thus it can be interpreted as social influence not having positive relationship with intention to purchase. Whereas from the t-statistics value of 1.781, it can be concluded that t-statistics < 1.96. Based on this result, it can be concluded that the eight hypothesis is rejected since social influence has no significant impact on intention to purchase in online shopping at KKN Café in Sidoarjo and Surabaya. This is supported by (Rotib et al., 2021) that stated that perceived ease of use has no impact on intention to use. H8: Social influence has impact on intention to purchase with mediation of attitude to use in online market place. The coefficient value of the impact of peer influence on intention to purchase is -0.029 thus it can be interpreted as peer influence not having positive relationship with intention to purchase. Whereas from the t-statistics value of 0.256, it can be concluded that t-statistics < 1.96. Based on this result, it can be concluded that the ninth hypothesis is rejected since peer influence has no significant impact on intention to purchase in online shopping at KKN Café in Sidoarjo and Surabaya. This is supported by (Jeng & Tzeng, 2012) that stated that social influence will never impact Intention to Use. H9: Peer influence has impact on intention to purchase with mediation of attitude towards to use in online market place.

Managerial Implication
1. This research can help companies and other researchers to find out more about the relationship between TAM factors and social factors on attitude towards to use and intention to purchase of customers in shopping using online applications. 2. After reading this research, companies and other researchers can find out the factors in TAM factors and Social factors that can affect attitude and intention to purchase. According to the results of this study, usability and ease of use, and peer influence are factors that affect attitude towards mobile app use. 3. After knowing that TAM factors and social factors have no effect on intention to purchase, what needs to be done as a suggestion is to strengthen the products or services offered by KKN Café. perceived usefulness has the greatest impact on attitude, so companies need to pay great attention to the ease of use of the application, especially for novice users and regular users of the application. 4. Online business managers can observe that the characteristics of perceived usefulness, perceived ease of use, and peer influence are able to create positive attitude Towards to Use, but the characteristics of perceived usefulness, perceived ease of use, peer influence and attitude towards Mobile App Use need to be improved in order to be able to create shopping intentions. 5. Based on the findings in this study, it was found that things related to TAM Factors have higher values than those related to social factors. This can be a suggestion for KKN Café managers to strengthen the usability and convenience of the application for implication towards attitude towards to use and intention to purchase.

CONCLUSION
The conclusion of this research shows that: 1. Perceived usefulness has impact on attitude towards to use and has the highest impact on TAM Factors. Thus, the first hypothesis (H1) is accepted. 2. Perceived ease of use has impact on attitude towards to use. Thus, the second hypothesis (H2) is accepted. 3. Dalam social factors, social influence has no impact on attitude towards to use. Thus, the third hypothesis (H3) is rejected. 4. Peer influence has impact on attitude towards to app use. Thus, the fourth hypothesis (H4) is accepted. 5. Attitude towards to use has no impact on intention to purchase. Thus, the fifth hypothesis (H5) is rejected. 6. In TAM Factors, perceived usefulness and perceived ease of use with mediation of attitude towards to use has no impact on intention to purchase. Thus, the sixth hypothesis (H6) and seventh hypothesis (H7) are rejected. 7. In social factors, social influence with mediation of attitude towards to use has no impact on intention to purchase. Thus, it can be concluded that the eight hypothesis (H8) is rejected.
8. Lastly, in social factors is peer influence with mediation of attitude towards to use has no impact on intention to purchase. Thus, the ninth hypothesis (H9) is rejected.
Based on the results of this research, several suggestions are proposed as follows: 1. For manager of KKN Café: M of KKN Café is recommended to strengthen the usability and convenience because the variables perceived usefulness and perceived ease of use have significant influence on attitude towards to use. In addition, it is important for business managers to create the best impression for each individual because peer influence variable has a significant effect on attitude towards to use, whereby the influence of peers can determine the attitude of the individual. Online business managers in particular must optimize the use of online applications to promote and increase product sales. 2. For further research: a. For future researchers who will conduct similar research, it is necessary to conduct further research on several factors other than the variables in this research (perceived usefulness, perceived ease of use, social influence, and peer influence) that have impact on intention to purchase through attitude towards mobile app use. b. Future researchers is expected to be able to examine the scope of a population that is different from and wider than the research population in this study.