For this study, we selected 80 second- and third-year undergraduate students of stomatology at Lanzhou University. The random setting as follow: total of 166 students were divided into two groups (male/female) firstly according to gender. Then 40 students were randomly selected in each of two packets using a random-number table consisted of 80 participants ultimately, which can draw random sampling scientifically. The subjects were randomly divided into four groups (n=20) with a similar male-to-female ratio following CONSORT guidelines. The four groups included one control group and three experimental groups: (1) jaw simulation model as control, (2) virtual simulation, (3) virtual-jaw (V-J), (4) jaw-virtual (J-V). The age distribution of the participants was in the range of 18 to 21 years with no regional differences. None of the subjects had received oral implantology courses before this study. Informed consent was obtained from all participants and the study protocol was approved by the Ethics Committee, School of Stomatology, Lanzhou University.
2.2. Study procedure
2.2.1 Theory teaching of implantology
Before receiving the operation training, all students participated in the theory courses and underwent initial assessment (See Fig. 1 for the flow chart). Each lecture was two hours long and taught by instructors with more than 10 years of clinical experience. The theoretical content covered implant apparatus, preoperative preparation, basic principles of implant, applied anatomy, and operative procedure, and based on Oral Implantology, second edition . The first theory exam was a multiple-choice questionnaire taken by all participants.
2.2.2 Implanting operation teaching
The operation training lasted eight hours over four days. The site of operation was the mandibular molars recreated on a virtual simulation system (Zhonghui Technology Institute, UniDental-MS01, China) and jaw simulation model (Fig. 2 A, B). The implant systems followed the Dentin implant surgical operation manual  and each group used a dental implant toolbox (Fig. 2 C). The order of the V-J and J-V groups was the opposite, that is, the eight-hour long operation was split into two four-hour operations for each group. This way, we evaluated the operational training sequence on effect of teaching.
2.2.3 Preparation before examination
The mandible of a pig (Yorkshire) was scanned using cone-beam computed tomography (CBCT). The imagine data were imported into a software (Sirona Dental, GALILEOS Viewer, German). According to the bone condition and anatomical structure, the best implantation site on the mandibular second molar was designed by the teachers (Fig. 3 A). Subsequently, the ideal site was assessed for deviation from the actual site (Fig. 3 B).
2.2.4 Operational assessment and secondary examination
In the participants’ assessment, the second mandibular molars of a pig were extracted as the implant site. The assessment was double-blinded and evaluated by professional dentists, it included two parts: (1) a subjective evaluation with students’ “preoperative preparation” and “intraoperative operation” procedures (See Table 1 for details on the evaluation); (2) an objective evaluation of the images obtained using the CBCT. The former part focused on the standardization and professionalism of the operations, while the latter part was used to judge the effects of implanting. Then, a second theoretical test was conducted to analyze the differences between the first and second testing scores and, thus, determine whether operation training could improve students’ theoretical knowledge.
2.2.5 Implant accuracy
The CBCT scan of the implanted site and the imported data were used to evaluate the rationality of implanting in order to calculate the angular, apical, and shoulder deviations. (Fig. 4)
2.2.6 Questionnaire survey
The degree of satisfaction from the students was gathered through the questionnaire using a Likert scale after the teaching session . Each item was rated, with 5 indicating “strongly agree,” 4 indicating “agree,” 3 indicating “neither agree nor disagree,” 2 indicating “disagree,” and 1 indicating “strongly disagree.”
2.2.7 Statistics analysis
a comparison between the four groups was done through an analysis of variance (ANOVA) in case the data conformed to normal distribution with homogeneous variance. For data not conforming to the normal distribution, a non-parameter test was adopted. The study compared the theoretical scores before and after the operation through a paired sample t-test. The operational assessment was based on a one-way ANOVA.