Deriving the Pricing Power of Product Features by Mining Consumer Reviews
Increasingly, user-generated product reviews serve as a valuable source of information for customers making product choices online. The existing literature typically incorporates the impact of product reviews on sales based on numeric variables representing the valence and volume of reviews. In this paper, we posit that the information embedded in product reviews cannot be captured by a single scalar value. Rather, we argue that product reviews are multifaceted, and hence the textual content of product reviews is an important determinant of consumers' choices, over and above the valence and volume of reviews. To demonstrate this, we use text mining to incorporate review text in a consumer choice model by decomposing textual reviews into segments describing different product features. We estimate our model based on a unique data set from Amazon containing sales data and consumer review data for two different groups of products (digital cameras and camcorders) over a 15-month period. We alleviate the problems of data sparsity and of omitted variables by providing two experimental techniques: clustering rare textual opinions based on pointwise mutual information and using externally imposed review semantics. This paper demonstrates how textual data can be used to learn consumers' relative preferences for different product features and also how text can be used for predictive modeling of future changes in sales. This paper was accepted by Ramayya Krishnan, information systems.
Methods to increase response rates to postal questionnaires.
BACKGROUND Postal questionnaires are widely used for data collection in epidemiological studies but non-response reduces the effective sample size and can introduce bias. Finding ways to increase response rates to postal questionnaires would improve the quality of health research. OBJECTIVES To identify effective strategies to increase response rates to postal questionnaires. SEARCH STRATEGY We aimed to find all randomised controlled trials of strategies to increase response rates to postal questionnaires. We searched 14 electronic databases to February 2003 and manually searched the reference lists of relevant trials and reviews, and all issues of two journals. We contacted the authors of all trials or reviews to ask about unpublished trials. Where necessary, authors were also contacted to confirm methods of allocation used and to clarify results presented. We assessed the eligibility of each trial using pre-defined criteria. SELECTION CRITERIA Randomised controlled trials of methods to increase response rates to postal questionnaires. DATA COLLECTION AND ANALYSIS We extracted data on the trial participants, the intervention, the number randomised to intervention and comparison groups and allocation concealment. For each strategy, we estimated pooled odds ratios and 95% confidence intervals in a random effects model. Evidence for selection bias was assessed using Egger's weighted regression method and Begg's rank correlation test and funnel plot. Heterogeneity among trial odds ratios was assessed using a chi-square test at a 5% significance level and the degree of inconsistency between trial results was quantified using I(2). MAIN RESULTS We found 372 eligible trials. The trials evaluated 98 different ways of increasing response rates to postal questionnaires and for 62 of these the combined trials included over 1,000 participants. We found substantial heterogeneity among trial results in half of the strategies. The odds of response were at least doubled using monetary incentives (odds ratio 1.99, 95% CI 1.81 to 2.18; heterogeneity p<0.00001, I(2)=78%), recorded delivery (2.04, 1.60 to 2.61; p=0.0004, I(2)=69%), a teaser on the envelope - e.g. a comment suggesting to participants that they may benefit if they open it (3.08, 1.27 to 7.44) and a more interesting questionnaire topic (2.44, 1.99 to 3.01; p=0.74, I(2)=0%). The odds of response were substantially higher with pre-notification (1.50, 1.29 to 1.74; p<0.00001, I(2)=90%), follow-up contact (1.44, 1.25 to 1.65; p<0.0001, I(2)=68%), unconditional incentives (1.61, 1.27 to 2.04; p<0.00001, I(2)=91%), shorter questionnaires (1.73, 1.47 to 2.03; p<0.00001, I(2)=93%), providing a second copy of the questionnaire at follow-up (1.51, 1.13 to 2.00; p<0.00001, I(2)=83%), mentioning an obligation to respond (1.61, 1.16 to 2.22; p=0.98, I(2)=0%) and university sponsorship (1.32, 1.13 to 1.54; p<0.00001, I(2)=83%). The odds of response were also increased with non-monetary incentives (1.13, 1.07 to 1.21; p<0.00001, I(2)=71%), personalised questionnaires (1.16, 1.07 to 1.26; p<0.00001, I(2)=67%), use of coloured as opposed to blue or black ink (1.39, 1.16 to 1.67), use of stamped return envelopes as opposed to franked return envelopes (1.29, 1.18 to 1.42; p<0.00001, I(2)=72%), an assurance of confidentiality (1.33, 1.24 to 1.42) and first class outward mailing (1.12, 1.02 to 1.23). The odds of response were reduced when the questionnaire included questions of a sensitive nature (0.94, 0.88 to 1.00; p=0.51, I(2)=0%), when questionnaires began with the most general questions (0.80, 0.67 to 0.96), or when participants were offered the opportunity to opt out of the study (0.76, 0.65 to 0.89; p=0.46, I(2)=0%). AUTHORS' CONCLUSIONS Health researchers using postal questionnaires can increase response rates using the strategies shown to be effective in this systematic review.
Rational Ignorance at the Patent Office
It is common to assert that the Patent and Trademark Office does a bad job of examining patents, and that it should spend more time and money weeding out bad patents. In this article, Professor Lemley challenges that conventional wisdom. Using available data regarding the cost and incidence of patent prosecution, litigation, licensing and other uses of patents, he demonstrates that strengthening the examination process is not cost effective. The core insight is that very few patents are actually litigated or licensed; most simply sit on a shelf unused, or are used only for noncontroversial purposes like financing. Because of this, society would be better off spending its resources in a more searching judicial inquiry into validity in those few cases in which it matters than paying for a more protracted examination of all patents ex ante. In economic terms, the patent office is "rationally ignorant" of the objective validity of the patents it issues.
time series decision making reinforcement learning mathematical model decision support system embedded system magnetic resonance magnetic resonance imaging united kingdom breast cancer developing country resonance imaging motion planning confidence interval humanoid robot social science experimental design linear algebra nash equilibrium data assimilation standard deviation black box autoregressive model functional magnetic resonance interest rate exchange rate design of experiments process capability prostate cancer fuzzy environment functional magnetic triangular fuzzy number rate model citizen science scientific publication open-source software choice behavior econometric model programming paradigm ordinary least square business cycle prefrontal cortex tumor cell process capability indice evaluation procedure capability indice risk aversion breast cancer patient hearing impairment mental disorder digital object identifier humanoid robotic demand function federal reserve trade credit aggregate datum sample variance major depressive disorder monetary policy entropy maximization major depressive circulating tumor cell circulating tumor mobile money cross-sectional datum marginal model rejection sampling policy rule cell nucleu financial cost mood disorder interpretation proces real interest rate trust game institutional investor nucleus accumben learning disorder money demand metastatic breast accidental fall positive theory genetic heterogeneity review [publication type] word list by frequency neuritis, autoimmune, experimental projections and prediction occur (action) departure - action job stream value (ethics) exhibits as topic reuse (action) contain (action) cellular targeting null value advance directive - proxy benchmark (computing) collections (publication) coping behavior deny (action) hearing loss, high-frequency revision procedure rem sleep behavior disorder codependency (psychology) pervasive informatic amygdaloid structure sensorineural hearing loss (disorder) norm (social) contingency (philosophy) behaviorial habit choose (action) persistence (computer science) eighty nine neoplasm metastasi columbia (supercomputer) expectation–maximization algorithm relocation of home or business manufactured supply mediator brand of benfluorex hydrochloride public-private sector partnership academia (organization) document completion status - documented entity–relationship model sex characteristic uc browser mike lesser the australian nat friedman lars bak (computer programmer) research support as topic lars gene substance-related disorder expected utility hypothesi fama im economic complexity index metastatic carcinoma conflict (psychology) rule (guideline) kind of quantity - equilibrium john d. wiley baseline (configuration management) transaction counts and value totals - contract contract agreement standards characteristic file spanning ephrin type-b receptor 1, human published comment copy (object) arabic numeral 0 united state