1 很多理论必须要建立在大量调查数据之上
2 科学家在进行真正实验前会应用想象实验，如果结果很清楚就没有必要进行真正实验了。同时很多理论无法收集到数据，只能靠假设。
3 过分依靠数据也会使理论僵化。
What is the correct procedure of making a theory, and what kind of theory are we going to have? In response to this issue, we need to analysis from a dialectic perspective. The author declared that we need data to obtain the accuracy of our theory. However, it simplified this issue. The final judgement, in my point of view, should depend on the nature of our investigations and the truth of concepts.
Most researches and theories, as in the realm of economics,science as well as social, it is true that we can not make the final conclusion without the case-by-case data, and the source of these data, must be identified as a matter of importance. As the first and lowest level in the process of making a theory, data itself do not carry any meaning, what really counts for is the interpretion and analysis based on these data. Then the data themselves turn out to be a group of information that represent the quantitative and qualitative of a variable or a set of a variable, rather than a result of measurements. What's more, the quality of the data, their quantities-- whether they were sampled randomly, or how they measured--may have an implication on the effectiveness and dependability of the research. To sum, a theory can not be created without data, and how the data was constructed and defined may have an implication on the quality of theory.
Of course, not every realm of question was eager for data, like arts and humanities, the priori investigations changed to be "thought experiments", rather than emperical experiments. Scientists tend to use thought experiments before the real experiments started, if the results of the thought experiments will be so clear that there will be no need to conduct a physical experiment at all. Besides, there are also cases where these particular physical experiments are impossible to conduct, the common goal of thought experiments are to explore the potential consequences of the principle in question. Among scientists , Galileo and Einstein were the most impressive thought experimenters, but thought experiments existed throughout the middle ages. One of the most beautiful examples of the thought experiment was to show that space is infinite:"If there is a purported boundary to the universe, we can toss a spear at it. If the spear flies through, it isn't a boundary after all; if the spear bounces back, then there must be something beyond the supposed edge of space, a cosmic wall that stopped the spear, a wall that is itself in space. Either way, there is no edge of the universe; space is infinite".
However, we must not over-emphases on the usage and dependability of the data, that is the symble of doctrinairism and bookishness. And over emphases of data before making a theory will certainly hamper the development of innovation theories. The theories are based on assumptions, if we do not establish our research until the data is prepared, the enthusiasm and creativity within the scientists will be hind. Hawking's theory, if according to the author, will be totally unreasonable and a "grave mistake". As a matter of fact,
current research conditions are still inadequate for gathering data to support Hawking's theory, which is wildly acknowledged in scientific world. So, data is not everything in creating a theory.
To sum up, data is indispensable in many realm of investigations and researches, at the same time, we need a balance when using them to creat a theory. In addition, thought experiment also displace its value in the arts and humanities. Thus, the author's claim was one-sided. There are many ways to make a theory, this complex issue depends on the nature of the researches and our diversification of thoughts.