Market Research Summart

...

方法：Secondary data analyzed in a quantitative manner; surveys; panel; observation

Cross-Sectional Design(一个东西的几个时间点)/Longitudinal Design（sample survey在连续的几个时间点）

Continuous panels: consumer purchase panels, unrepresentative samples, expensive, atypical, expensive

Causal Research：know which variables are the cause (independent variables) and which variables are the effect (dependent variables) of a phenomenon. METHOD: Experiments

Guidelines：

A questionnaire must be clear, unambiguous and uniformly workable.

The design must minimize errors from respondents, interviewers and coders.

It should arouse respondents interest in eliciting answers that are as close an possible to the truth.

Questions should be held to be not more than 20 words if at all possible.

Keep in mind the objective of the research.

Think in detail. All questions asked must be aimed at providing the information necessary for making good decisions.

Irrelevant questions should not be asked.

5 scale

Nominal: For categorizing only, resulting in classes which may be given an arbitrary numeric designation.（分类，数字随便定义）；No order or distance relationship.（没有顺序距离关系）No arithmetic origin.（数字没含义）No generally used measure of dispersion.（分散程度没有关系）Only a limited number of statistics, all of which are based on frequency counts, are permissible,

例, percentages, and mode. 男女性别；季节出生；以下课程上了哪门

Ordinal scale

123等级排列，级别距离等差别不同，等级不准确

Any monotonic transformation will preserve order.

ordinal scales permit the use of statistics based on centiles, e.g., percentile, quartile, median.

例：比如地震等级，赛马顺序，品牌喜爱顺序；每天花多少时间上学，每周吸多少烟（选项间距不相等，比如最后一项>30mins 如果相同则为interval／1包，1包）；满意度（满意一点点，不满意，很满意）；level of education ； change in awareness score

Interval scale

间距相同equally spaced points on the scale；No natural origin; 通常>5

Ratios of values are not meaningful. Can be applied to nominal and ordinal data.

例 经纬度, IQ;

Ratio scale

拥有a natural or absolute zero.数字具有本身的含义

Examples: weight, height, number of children, waiting time. Income;temperature scale

[pic 6]

scale methods: ranking, rating, running prompt(一生来过多少次迪士尼，最近一年以及月), direct judgment (品牌贵便宜自己横线上截取)；Fractionation（比如品牌A甜度标准为1，同它相比别的牌子多少度量化）；constant sum（总分10，5个品牌分配分数）

Likert Summated Scale;分数来表示态度

semantic differential is a seven-point rating scale with end points associated with bipolar adjectives labels. 两级选择，powerful 3 2 1 0 -1 -2 -3 weak

Semantic Differential Profile (Snake Diagram) 连点画图

其他注意事项：7最多2最少；偶数个，奇数会偏向中间

6 qualitative studies

Focus Group Interview：unstructured-indirect.

缺点：Sample not representative e.g. close relationship affinity groups. Members are able to hear what each other person says.

In-depth Interview: 半／结构化；30mins – 1h； 一对一； exploratory/ pilot studies

to surface underlying motives, beliefs, and attitudes(为什么说，tell me more?)

Projective Techniques

应用： 1. 个人敏感社会倾向 2. 直接询问尴尬 3.Helpful when underlying motivations, beliefs, and attitudes are operating at a subconscious level.

The Third Person Technique：Respondents are asked for the view of 第三人.He will indirectly project his own feelings onto the third person.你怕飞吗？你觉得你邻居怕飞吗？

Word Association Tests ：test words

Sentence Completion Tests.

Story completion ：respondents are given part of a story – enough to direct attention to a particular topic but not to hint at the ending.

Cartoon tests： 卡通内容填写

7. descriptive research

Personal Interview（Interviewee biases）

Biases: 原因：social interaction between the interviewer and the interviewee

Reactive effect ：choosing to be a respondent affects his behavior；error occurs even when the subject is well intentioned and cooperative

Role selection bias：Respondents assume a different role in their research.

Demand effect：推测研究者的意图回答

Inability and unwillingness： 不记得一周喝多少可乐；不愿意泄露隐私；Social desirability bias，应该做而不是实际做的

Solution: 1. Indirect questions 2. Randomized responses technique

Randomized responses technique： 例

interviewer biases：Experimenter effect;characteristics of the interviewer: