Shulou(Shulou.com)06/03 Report--

When the tester gets the test task, he or she needs to examine two basic situations. The first category is the situation of testers:

What is the tester's testing experience, rich or lacking?

What is the tester's industry experience, familiarity or understanding of the product under test?

What does the tester know about the requirements of the product under test, familiarity or understanding?

The second category is the situation of the products under test:

What is the current stage of product development?

Is the product tested, what types of tests are used, and what features and attributes are covered?

What are the current risks or potential problems of the product?

Testers should carefully analyze and understand these situations. Under the pressure of time and business quality, testers need to drive testing activities according to the correct information in order to achieve better results.

First of all, testers need to be very clear about their own situation, that is, their own knowledge (Knowledge), including product knowledge, industry knowledge, testing technology, development technology and computer foundation.

The purpose of checking this knowledge is to quickly and effectively determine whether the problem found is a poor review defect during testing. That is, testers need to integrate all kinds of knowledge to construct a group of test prophets, so as to identify product defects efficiently. One possible way to construct test prophets is to refer to HICCUPPS (F) heuristic rules [Bolton05], which identify product defects through conformance checks.

N history (History): whether the current version of the product is consistent with the previous version.

N Image: whether the product is consistent with the vision of the development organization.

N similar products (Comparable Products): whether the products are consistent with similar products.

N statement (Claims): whether the product is consistent with the statement of the important person.

N user expectations (User's Expectations): whether the product is consistent with user expectations.

N the product itself (Product): whether the comparable functions in the product are consistent.

N purpose (Purpose): whether the product is consistent with its (express or implied) purpose.

N regulations (Statutes): whether the product is consistent with the applicable law.

N common sense (Familiarity): whether the product is consistent with common sense.

Although there are many ways to enrich the test prophet, it is difficult to achieve perfection without paying enough time cost (all expected outputs are clearly defined before testing). Therefore, we often encounter the following problems in the exploratory testing process:

The absence of a test prophet enables testers to determine in advance that the observed system behavior must be right or wrong.

N there is no single test prophet to show that a function works correctly at any time or under any circumstances.

N some functions seem to work properly, but they will fail in some cases and affect the correctness and applicability of other test prophets.

It can be seen that testers must encounter a lot of difficulties when constructing test prophets, so how to solve them? Here are three possible ideas.

N ignore this problem (perhaps the value of this information is not worth it from a cost perspective).

N simplify the problem (improve testability to get more information, analysis requirements, specifications, and code for simpler inspection rules).

N transfer the problem (consider the relevance of the problem and start with similar problems).

During the test design process, testers can also use heuristic methods (Heuristics) to generate more and better test ideas, such as [Bach21]:

N-guided word heuristics (Guideword Heuristics): some words or tags can guide testers to explore their own knowledge and experience and generate new test ideas.

N-trigger heuristic (Trigger Heuristics): ideas that exist in events or conditions can remind testers to experiment in another way, like an alarm clock for thinking.

N-subtitle heuristic (Subtitle Heuristics): it can help testers reconstruct test ideas and get more choice points.

N heuristic model (Heuristics Model): a set of systematic ideas can help testers control, manage, and mine more ideas.

In fact, people often use heuristic methods in their daily life. For example, we often use the following simple rules to deal with complex practical problems:

N drunk driving is very dangerous.

A bird in the hand is better than two in the bush.

N do not take high risks, can not get high returns.

People sometimes hide their passwords next to their computers and try to find them from there.

N Holiday shops usually open late.

If your computer behaves inexplicably, restart. If there is still a problem, reinstall the operating system.

If this is a really important task, your boss will follow it, otherwise, you can ignore it.

From these examples, we can see that the heuristic method is an empirical method, which puts forward a simple and more likely successful solution to complex problems. Using heuristic methods, people can quickly take action to explore answers in practice, thus avoiding endless problem analysis and making no progress. However, the heuristic method is only a "shortcut". It does not guarantee the "best answer", nor does it guarantee that the "right answer" is always provided. Therefore, a wise tester will not rely on a specific, single heuristic method, he will use a combination of heuristic methods, and adjust the test method and test prophet according to the results of practice.

This article is excerpted from the book "the Road to exploratory testing practice"

Shi Liang, Gao Xiang

Published by Electronic Industry Press

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.