In the new century of the knowledge-based economy, today's teachers have to face various challenges - the imperative of educational reform, the complexity of learning outcomes and the rapid social changes.

In facing the various challenges, an efficient teacher will be no longer satisfied with using the "instructional guide" as a sole indicator for instructional planning. Instead, she would have to adopt innovated methods to teach students and evaluate their progress according to the current "Zeitgeist" and the need of future world. A student must learn how to adjust to the existing reality, prepare for a future society, and create a scenario in order to achieve a real success. Thus, the instructional method and learning style have to be aligned to the new demand. Teaching and learning of knowledge and skills are only valued if they are geared towards this end, i.e., directed to students' future success. It is obvious that the score or grade on a student's record is not the only criterion of a student's achievement.

Recently, there have been many new insights on the research of intelligence. Not only has the concept of intelligence been extended but also the traditional assessment of intelligence has been challenged extensively. The new insights have provided educators wider and deeper dimensions for research as well as for instructional and curriculum design. Among these studies, two models deserve special attention. One is Harvard University Professor Howard Gardner's (1983, 1993) Theory of Multiple Intelligences (MI). The other is Yale University Professor Robert J. Sternberg's (1985, 1988) Triarchic Theory of Intelligence and his concept of Successful Intelligence (Sternberg, 1996). They both claimed that the intelligence assessed by traditional IQ tests could not portray the complete intelligence itself and is only related to the ability of academic achievement. It cannot predict the achievement of future career and the fulfillment of life. Gardner and Sternberg both have been seeking the "useful intelligence" in daily life. Gardner (1983) constructed seven components of intelligence, including linguistic ability, musical ability, logical-mathematical ability, bodily kinesthetic ability, spatial ability, interpersonal ability, and intrapersonal ability (the last two are called personal intelligence). The 8^th intelligence, the "naturalist intelligence", which has been added recently (Gardner 1999).

Sternberg (1985, 1988), defined intelligence from three aspects--componential, ontextual, and experiential. In the book, Successful Intelligence, Sternberg (1996) indicated that successful intelligence (analytical intelligence). Both Gardner and Sternberg conceived intelligence as multiple dimensional and flexible in nature.

Integrating Gardner and Sternberg's constructs of intelligence, the author has attempted to propose a construct to show the relationship between a successful career and a successful intelligence, in which personal intelligence is the core component (Figure 1).

Figure 1 illustrates the following:

1. There are four types of intelligence that lead to a successful career, i.e., the traditional intelligence (analytical intelligence) and the other three non-academic intelligences -personal intelligence, practical intelligence, and creative intelligence. A balanced development among these four types of intelligence can make a successful career and fulfillment of life.
2. There is a close relationship between academic intelligence and achievement in different academic subjects. Different kinds of academic intelligence can influence achievement in different subjects, such as linguistic intelligence for literature, logical-mathematical intelligence for math and science, musical intelligence for music, spatial intelligence for fine arts, and bodily-kinesthetic intelligence for sports, dancing, and drama.
3. Personal intelligence could be referred to as social intelligence or emotional intelligence, which involves intrapersonal intelligence and interpersonal intelligence.
4. Practical intelligence belongs to the field of cognition. It is an ability to apply knowledge to daily life or problem solving.
5. Creative intelligence includes creative thinking (cognition) and creative attitude (feeling). Sternberg (1996) believes that creative intelligence and practical intelligence have more impact on career achievement than does academic intelligence (traditional intelligence). Creative intelligence can be integrated with the executive ability of practical intelligence to produce stronger power or concrete effect, such as a creative problem-solving.
6. Academic intelligence is the basic requirement of a successful career, but it is not sufficient. A successful career (a successful job and contented life) requires a balanced development of the above four types of intelligence. On the other hand, personal intelligence is the core of achieving a successful career and the catalyst for the other constructs of intelligence.

It is believed that the MI model has its universal meaning. However, when it is applied to appraisal and teaching, cultural and social factors have to be taken into account. In other word, it has to be studied, experimented, and evaluated locally. That is why the author has been leading a research group in carrying out a 4-year joint research project (1999-2003) focusing on the development of MI (in the name of the development of multiple talents, DMT), under the sponsorship of the National Science Council, R.O.C. (Taiwan). The DMT model (refer to Figure 2) proposed for this study has three dimensions: Form of talents (11 forms, mainly based on Gardner's theory of multiple intelligences), function of talents (3 functions, based on Sternberg's conception of successful intelligence), and developmental stage (4 stages, from pre-school to senior high).

For example, in the study of Personal Intelligence (PI), the Personal Intelligence Inventory (PII) have been developed based on Thorndike's (1920) classical social intelligence, Gardner's (1983) conception of personal intelligence in his MI theory, Goleman's (1995) EQ, and Chinese cultural point of view. The 72 items of PII are all situation-oriented. There are three forms of PII: Form A is in the form of multiple choice, Form B is an open-ended questionnaire, Form C is a very individual experience-based questionnaire. All forms measure the same construct of personal intelligence (three domains, with 4 sub-scales for each domain) with the same criteria, using a 4-point rating scale. Based on the sample of 620 grade 5 and 6 pupils in the Taipei area, the reliability of the PII was satisfactory. The inter-domain and the inter-subscale correlation coefficients were high and significant. The test-retest reliabilities of the total scale for Form A was .80, whiles the three domains were .61, .61, and .75, respectively. The inter-rater reliabilities for the total scale and domain scales of Forms B and C were also around .80. However, it is somewhat affected by "social desirability". The correlation coefficients between the PII and the Social Desirability Scale were .29 (p<.05) for the total PII score and .36 (p<.01), .15 (p>.05), and .28 (p<.05) for the three PII domain scores (Intrapersonal, Interpersonal, and Interactive), respectively.

Using the PII, Wu and Chien (2000) found some personal intelligence differences between gifted and regular pupils. However, gifted group is not as superior as on academic performance to the regular one. Details are as the following:

Wu and Chien (2000) used the PII to assess the 5^th and 6^th elementary school gifted and regular pupils (total N=620) in the Taipei area. It was found that intelligence and gender did have significant effects on personal intelligence. The multivariate analysis showed significant Wilk's Lambda values (Ls), .946 (p<.01) and .932 (p<.01), respectively, for both intelligence and gender variables; there were no significant interactions. In terms of intelligence group, the gifted students' interpersonal ability ("guiding others", in particular) was better than regular ones (Table 1); in terms of gender factor, girls' intrapersonal abilities ("self-retrospection", in particular) and interpersonal abilities ("being respectful" and "being amiable to", in particular) surpassed boys', while boys only showed better than girls in "being humorous" in the interactive dimension (Table 2). It is interesting to note that it seems that gifted group, although in general is in a better position, is not as superior as on academic performance to the regular one.

In the second two years, the guidelines and principal contents of the Personal Intelligence-focused Multiple Intelligences Program (PI-MI program) for elementary school and junior high school are designed. The PI-MI program has been field-tested in an elementary school in Taipei (Wu & Chien, 2001). Based on the framework of MI, the PI-MI curricula under nine themes were designed. A pre-post equivalent group experimental design was adopted and carried out. The subjects comprised of 132 fifth grade students, 66 were gifted and 66 were regular. Among them half were assigned into experimental group (EG) and the other half the control group (CG). Males and females are even in number in each sub-group. The PII and the "School Life Inventory (SLI)" were pre- and post-administered to all subjects to assess the effect of the experiment. The EG students participated in a eight-week class-based teaching of PI-focused MI program in related courses according to regular school schedule, while the CG students received regular teaching. There were 6 teachers in gifted education involved in the PI/MI curriculum design and EG group instruction. They had been participating in the prime researcher-conducted special training program before and during the experiment. By means of 2x2 ANOVA and one-way ANCOVA, the collected data were analyzed. The main findings were as the following: (1) In the EG, both gifted and regular students' PI scores were significantly increased after the experiment; (2) Using pre-test score as a co-variate, the EG-CG comparisons showed significant experimental effect, especially in the "self-awareness" and the "conflict resolution" sub-scales; (3) In terms of school adjustment as assessed by SLI for EG, both gifted and regular students' negative coping behaviors (avoidance and blaming) were significant decreased after the experiment. In conclusion, the PI/MI program had positive effects on students' personal intelligence and coping behaviors (Wu & Chien, 2001). Please refer to Table 3 and Table 4.

There is a Chinese saying that "Every one's potential is useful in some way." No matter how smart or stupid a student is, he or she should have his/her merits and disadvantages along with different learning needs. A teacher must learn how to discover their students' diverse characteristics and needs and how to teach diversity through variety in order to realize an ideal of "school without failure". This is the goal of educational reform and, of course, a big challenge. The theory of multiple intelligences (talents) and the practice of multiple (flexible) instructional methods may be the key of resolution. We anticipate that the MI model will have to be incorporated into a well established educational reform, either teaching for MI, teaching by MI, or teaching about MI, so as to bring out a new era - not only "educational for all," but "excellence education for all.