Industry insiders believe that embracing big data and AI will become an important direction for the future development of the catering industry. It is also an effective method to save costs and improve service quality. A fast food company in Jinan went online to serve customers via a WeChat program during the epidemic. The move received a surprising welcome from nearby white collars, who no longer need to wait for food inside the eatery. "We were inspired and planned to set up thermal cabinets in some office buildings. Customers can order food in advance and we will deliver the food to the cabinets," said Song Yefei with the company. Many restaurant chains in China have also expanded their service online and stepped up efforts to develop semi-cooked or ready-to-serve food, as well as promoted their products in shopping malls or via live streaming to lure cooped-up foodies. "We suffered huge losses amid the epidemic. But we are gradually recovering with more restaurants open these days. We stay confident in the future as the consumption market in China is very big," said Hou. As the epidemic recedes and work and production resumption gains momentum, Chinese consumers' buying appetite inhibited by the epidemic is rebounding.

3. Consider a customer survey in the formal stage, and decide whether you would use a personal survey, telephone survey, or self-administered survey (mail, e-mail, or online). Give reasons for your choice of contact method and describe the survey process. In your chosen method, what are the major error sources? And what is your sampling design? 1) Contact method, sampling method and survey process, and the reasons for your choice. (5 points)

2) According to the survey mentioned in the case, nearly 70 percent of the respondents claim a willingness to try dishes cooked by robot chefs. You wish to estimate with 95% confidence (Z=1.96) that the allowance for sampling error will not greater than 4%. What the required sample size should be in this case? (4 points) 3) Figure out at least three sources of research error and the approaches to minimize these errors respectively. (3 points)