Big Data Use in Business PDF

CONTENT Big Data Use 1. Understand market reaction and brand perception. Businesses can assess consumer reaction to product announcements, pricing changes, policy changes, and other moves. They can gauge the effect of promotions, advertising campaigns, and...

CONTENT Big Data Use 1. Understand market reaction and brand perception. Businesses can assess consumer reaction to product announcements, pricing changes, policy changes, and other moves. They can gauge the effect of promotions, advertising campaigns, and publicity programs. And they can compare sentiment associated with their brands to those of competitors 2. Identify key buying factors → why people buy the product you're selling Analysis of website click streams, product reviews, product registration information, and public reaction to products can help merchandisers and product planners understand what buyers seek and why they buy what they buy 3. Segment populations to customise actions. [see research paper] because not all customers are the same → diff interests, wants When marketers can understand the attributes of different segments of the market and how members of each segment react to products, advertising, and promotions, they can tune marketing actions to achieve the best results within each segment 4. Enable experimentation. When business leaders can quickly assess the effect of changes in products, marketing, and operations, they can perform experiments to help assess the value of changes and determine the best course 5. Predict outcomes. [see research paper] Insight into the result of previous changes coupled with better understanding of market reactions lets businesses better predict the results of changes like price increases, advertising, and promotions 6. Enable new business models. Information services that aggregate and analyse Big Data have grown up in industries like healthcare, real estate, finance, retail, and others Data on economics are very important to understand how different sectors of an economy are doing. Examples of big data collection/uses Loyalty programme/scheme ○ When you check out your items with your loyalty card, you are divulging two important pieces of information: (1) your demographic profile; and (2) shopping habits / patterns → data for supermarket ○ To serve up movies with advertisements based on what you bought from the supermarket (these are known as targeted ads) ○ To offer you vouchers for salad and fruits, if you usually purchase unhealthy stuff ○ To customise your log-in page so that it is easier to find the products their data suggests you will buy ○ To offer suggestions on what to purchase based on your demographic profile (gender, age, etc.) Payment using credit card ○ Credit cards anonymous but card numbers can still be tracked ○ How often this person visits the supermarket? How much does he buy each time? What does he buy, which branch does he visit? Does he have a pet dog / cat? What brand of toothpaste/shampoo/detergent/dishwashing liquid does he buy? Is he sensitive to special offers? Real time data analytics available due to ○ the Internet of Things; and ○ the extensive use of sensor data – predictive algorithms are used to determine when a particular unit will need servicing, or when replacement parts will need to be ordered. Tesco use real time data analytics ○ to continue to gain a better understanding of the changing nature of consumer behaviour ○ to create efficiencies in their logistics and distribution chains in order to keep down costs and minimise environmental impact ○ to face up to the challenge of emerging business models which compete with their own ○ to reduce the amount of food which goes to waste at their stores Sales forecast ○ Modelling trends in customer behaviour ○ Not just how they shop in each store – but how they shop for each product.  ○ Analytics ensures that the correct products are always in stock, and not going to waste. It ensures that inventory is managed in the most efficient manner Bluetooth beacons ○ small, unobtrusive electronic devices that are hidden throughout the grocery store. An app on your phone that communicates with them informed the company not only that you had entered the building, but that you had lingered for two minutes in front of the ice-cream freezer ○ They use little energy, and they work well indoors. That has made them popular among companies that want precise tracking inside a store ○ E.g management of malls can know high density areas → adjust rental price accordingly Counting Past: usually to assess the numbers available for military duty or taxation. Today, governments collect statistics on all sorts of things to aid planning to make decisions that are data-driven, e.g., the census provides an insight on the demographic profile of a country. From the census, we know that the population in Singapore is ageing rapidly, and this information can be used to project the number of hospital beds that are required. Measurement (through counting) is a legitimate and necessary activity in the study (and management) of society. RESEARCH PAPERS Paper 1 (e-commerce) Big data research Aim: To get a thorough understanding of the impact of images on the outcome of an online transaction Online marketplaces enable online shopping, which has become so popular that it now threatens traditional retail businesses Online shopping process 1. Recognise a need for some product or service 2. Search online marketplace for need-related information 3. Get attracted about certain products or services 4. Evaluate or compare alternatives 5. Make final decision for purchasing 6. Be provided with after-sales service Online transaction depends on 1. product (price & quality) 2. Advertising (listing title, description, image(s)) 3. Service (shipping, return policy, communication, security) 4. Buyers intentions 5. Sellers qualifications Online peer-to-peer marketplace characteristics ○ Vast – typically millions of products on sale, millions of sellers, millions of buyers, millions of transactions ○ Physical examination of the products is not possible ○ No face-to-face interaction, leading to: Product uncertainty Seller uncertainty Buyer uncertainty Buyers rely on 1. textual description 2. Images → Product images can be used to convey information to the potential buyer about the product (colour, quality, condition, size, …) Some products rely more on their visual appeal, e.g., clothing and jewellery. Other products, e.g., electronics and cameras, rely more on their specifications and functionality 3. reputation of players (i.e., crowd wisdom) 4. communication (typically through text) eBay ○ Diversity of products ○ Diversity of product conditions  ○ Diversity of sellers ○ Sellers have little or no photography skills, leading to a huge variation in the image quality, even for the same item. Data preparation Dataset 1 Dataset 2 Randomly chosen product listings from the year 2011–2012 Randomly sampled listings created on several Used to study the impact of display type or the use of multiple random selected months in the year of 2012 images on conversion rate and seller type Used to study the impact of image quality Results ○ Custom photos>stock photos: using an photo of the actual item increased the likelihood that it would be sold→ buyers what to see real thing ○ Larger photos (which provide more visual information for buyers) increase the likelihood of a sale→ note: there are 5 display types with different price to upload ○ By giving buyers more visual information from different views of the object, the probability of success is increased→ there is only a small impact beyond three images ○ A high quality, or professionally photographed image may help to draw attention to a product or boost the buyer’s confidence for a product Scoring image quality Based on the preference of product image from e-commerce perspective, a quality score that emphasises on the clarity of the foreground object, high brightness, strong contrast between foreground and background, clean and uniform background, etc., was calculated. The quality score ranged from 0 to 255, where a score of 0 refers to the poorest quality image, while a score of 255 refers to the best quality image. ○ Less frequent buyers (categories D and E buyers) prefer listings with high quality images which can provide more assurance on the product, whereas frequent or seasoned buyers show less concern on possible risk for listings with low quality images Buyer and seller categories ○ Seller Categories Sellers are divided into different levels (casual, bronze, silver, gold, platinum, titanium) based on the increasing amount of sales made and revenue generated. ○ Buyer Categories Buyers are also grouped into several levels (A, B, C, D, E) based on their spend over the last 12 months. A-type buyer spends the most, while E-type buyer spends the least. Resale items ○ When buyer re-lists item he/she bought from initial state within 3 months ○ The profit rate (𝑅𝑃) was calculated for the resale items (𝑃𝑏 refers to the resale price and 𝑃𝑎 refers to 𝑃𝑏−𝑃𝑎 the initial purchase price) → 𝑅𝑃 = 𝑃𝑎 Product watch ○ An item “added-to-watch-list” may not necessarily lead to a check out action. ○ Unlike actions such as “add-to-cart”, whereas users have to either confirm their decision for purchase or remove the item from the cart later on, “watch” action comes with lower cost and pressure since no further action is required. ○ The number of people “watching” an item is hidden from the buyer. ○ A “watch” signals a buyer’s heightened interest in an item Paper 2 (physical activity data) Big data The analysis of a large-scale measurement of physical activity patterns across 111 countries worldwide. The measurements were made using smartphones with onboard accelerometers for automatic recording of physical activity throughout the day. Found that physical activity is distributed unequally (i) between countries Japan has a lower activity inequality (a smaller spread of steps) compared to Saudi Arabia (ii) within countries (iii) between genders  ○ Activity in females is lower than activity in males (i.e., there is a “gender gap” in activity levels). ○ In countries with low activity and high activity inequality, the gender gap in activity is amplified. Found that the built environment, specifically the walkability of a city, impacts physical activity. Activity inequality is associated with higher obesity levels A country’s activity inequality is a better predictor of obesity prevalence compared to the average number of steps recorded Walkability of a city ○ modifiable extrinsic factor that could increase activity levels and reduce activity inequality and the gender activity gap ○ Higher walkability scores are associated with lower activity inequality. ○ Walkable environments increase physical activity during work and leisure times. Research limitations ○ Biassed towards: individuals of higher socioeconomic status (smartphones are expensive) people interested in their physical activity and health ○ Time spent in activities where it is impractical or impossible to carry a smartphone will not be captured ○ There are also systematic differences based on gender and age Paper 3 (dog-human companionship) Not big data Aim: Dog companionship has attitudinal, experiential, and behavioural components that underlie it, and it is our objective here to uncover its underlying dimensions. Data collection: ○ 749 dog owners who filled out a self-administered questionnaire, in three settings: mall-intercept setting (424 respondents) veterinarian waiting-room setting (219 respondents) dog-owning Internet discussion-group setting (106 respondents) ○ A convenience sample of dog owners was obtained. ○ Fifty-seven Likert-type questionnaire items Method of analysis : exploratory factor analysis ○ It is a useful tool for investigating variable relationships for complex concepts such as socioeconomic status, dietary patterns, or psychological scales (and dog companionship is a complex concept). ○ It allows researchers to investigate concepts that are not easily measured directly by collapsing (reducing) a large number of variables into a few interpretable underlying factors Seven Dimensions 1. symbiotic relationship The mutually beneficial bond between person and dog. A combination of enjoying the nurturing component of having a dog along with the benefits received by both parties. The human is happier, less stressed, less lonely, safer, and calmer. The dog is treated as a child/person who is fed, cared for, and psychologically nurtured. 2. dog-oriented self concept The importance of the dog(s) to the human’s self-concept and social self. The dog is both an extension of self and the human’s best friend. The person perhaps spends less time with other people in general due to her/ his relationship with her/his dog but seems to have a better relationship with other “dog people” and with those willing to accept the dog as a part of the owner. 3. anthropomorphism The dog is seen as more of a person and less of an animal. The dog is perceived as a child surrogate or as part of the family, who can be communicated with much like another human. Here, the dog owner has opportunities to learn from the dog. This dimension corresponds closely with the conceptualisation of the pet as child surrogate 4. activity/youth The increased activity levels of the person due to dog ownership, where the person feels young or like a kid and is more physically active.  Corresponds with “childlike and playfulness” and suggests that dogs may serve as the catalyst to remove people's inertia to physical activity and to make them more physically active. The most utilitarian aspect of pet ownership, in which humans receive the benefits of increased exercise and, presumably, better health. 5. boundaries The lack of limits imposed on the dog by her/his owners. This factor reflects the appropriateness of letting a dog have free run of the household or of not setting boundaries, such as allowing the dog to sleep on the bed or to get on the furniture. A higher score means that fewer boundaries are set. 6. specialty purchases The extent to which people are willing to make a special effort to acquire products for their dogs. Relates to the specialty status of both the shopping behaviours and the dog-related products that are purchased. Specialty products include brand name, foods, luxury items, home-cooked meals, and travel 7. willingness to adapt Refers to people’s readiness to change their patterns of living and consuming to accommodate their dogs. Choice of living space, home, or outdoor set-up and choice of vehicles are altered due to the presence of the dog. The dog owner who scores high on this dimension considers the dog in many household purchases and arrangements. Paper 4 (incompetence) Aim: explain why people imperfectly appraise themselves (miscalibration). ○ They usually arrive at inflated, overly-optimistic self-appraisals. ○ The incompetent grossly overestimate their skills and abilities In many domains in life, success and satisfaction depend on knowledge, wisdom, or savvy in knowing which rules to follow and which strategies to pursue. This is true for many tasks in the social and intellectual domains. People differ widely in the knowledge and strategies they apply in these domains, with varying levels of success. Some of the knowledge that people apply to their actions are sound, and meet with favourable results. Others are imperfect at best, and wrong-headed, incompetent, or dysfunctional at worst. When people are incompetent in the strategies they adopt to achieve success and satisfaction, they suffer a dual burden. Not only do they achieve erroneous conclusions and make unfortunate choices, but their incompetence robs them of the ability to realise it. Instead, they are left with the mistaken impression that they are doing just fine. Incompetent individuals lack metacognition, metamemory, metacomprehension, or self-monitoring skills – they have the ability to know how well one is performing, when one is likely to be accurate in judgement, and when one is likely to be in error. ○ Implication: The same knowledge that underlies the ability to produce correct judgement is also the knowledge that underlies the ability to recognise correct judgement → above-average effect: the tendency of the average person to believe he or she is above average Predictions 1. Incompetent individuals, compared with their more competent peers, will dramatically overestimate their ability and performance relative to objective criteria. 2. Incompetent individuals will suffer from deficient metacognitive skills, in that they will be less able than their more competent peers to recognise competence when they see it. 3. Incompetent individuals will be less able than their more competent peers to gain insight into their true level of performance by means of social comparison information. 4. The incompetent can gain insight about their shortcomings, but this comes by making them more competent, thus providing them the metacognitive skills necessary to be able to realise that they have performed poorly. Four studies Study 1: humour Study 2: logical reasoning Study 3: logical reasoning Study 4: english grammar -A skill which requires sophisticated -Presented questions - Participants were given - Complete a test knowledge and wisdom about the tastes which there is a definitive a test of logic based on assessing their and reactions of other people answer the Wason selection task, knowledge of the - At the end of the test, and asked to assess American Standard -Asked to assess joke quality on an participants were asked to themselves Written English 11-point Likert scale, ranging from 1 (not estimate how many of the - Half the participants - Then, rate their overall at all funny) to 11 (very funny) questions they had gotten were given a short ability to recognise correct - Finally, participants rated their “ability to right training session designed grammar, how their test recognise what’s funny” by providing a - Largely similar results to to improve their logical performance compared percentile ranking: 0 (“I’m at the very that of Study 1 reasoning skills→ training with that of their peers, bottom”) to 50 (“I’m exactly average”) to sessions would provide and how many items they 99 (“I’m at the very top”) incompetent individuals had answered correctly with the metacognitive on the test - The incompetent participants (those skills to evaluate which - Several weeks after whose test scores fall in the bottom test problems they had Study 4, the bottom and quartile) grossly overestimated their ability answered correctly and top quartile performers relative to their peers incorrectly, and reduce were invited back for a - Even participants in the bottom quarter miscalibration of their follow-up→ They were of the distribution tended to feel that they ability estimates given the tests of five of were better than the average - Finally, participants were their peers to grade Participants in the other quartiles did not asked to indicate which - After being shown the overestimate their ability to the same items they had answered tests of their peers, they degree correctly, and which they were asked to re-rate IN FACT, those in the top quartile had not, and to rate their their ability and underestimated their ability relative to their ability and test performance relative to peers performance once more their peers Problem: humour is subjective Problem with extremely competent individuals ○ Although they perform competently, they fail to realise that their proficiency is not necessarily shared by their peers Dunning-Kruger Effect ○ When incompetent people are too incompetent to realise they are incompetent Paper 5 (taxis) Big Data Yellow taxis have fewer accidents than blue taxis because yellow is more visible than blue → increase odds that other drivers would have sufficient response time to avoid to potential accident with a yellow taxi (colour conjecture) No difference in: ○ Far structure and monthly rent → no difference in economic incentive to motivate different driving behaviours ○ average speed between two colours→ difference almost fell in range of +/- 1km/h Argument: ○ The difference in the accident rates is not due to the colour of the taxis, but due to the drivers. ○ Perhaps the yellow taxi drivers underwent better training that emphasised passenger safety. It was the training that made them more careful, and this translated into the lower accident rates. ○ The taxi colour has little or nothing to do with the accident rates Additional variables not taken into account ○ reaction time measured using age of the driver as the proxy variable ○ eyesight measured using on three levels: (1) whether eyesight correction is used; (2) the “power” of the lenses; (3) any additional issues (e.g., astigmatism) ○ fatigue measured using the hours into the shift ○ weather condition at accident site measured using the categories provided by the Meteorological Service of Singapore ○ location of accident, measured using the geographic coordinates (longitude and latitude) ○ road configuration measured as a category (T-junction, slip road, etc) ○ time of day  measured using category provided by LTA/Singapore Police Force (peak vs non-peak), this is an indication of road congestion ○ driver experience, measured using the number of years of driving ○ taxi driver experience, measured using the number of years of driving a taxi (like pilots, who have the number of flying hours) ○ time category: (a) looking for passengers; (2) active taxi hire (passenger in taxi and on the way to destination); (c) taxi on call (taxi on the way to pick up a passenger who has made a booking) Paper 6 (tree density) Forest management: ○ Achieve environmental, social and economic goals effectively Security of forest and its inhabitants Preservation of ecosystem’s health Improvement of wood goods and services production ○ Traditionally national authorities control this process because woodlands usually cover a large extent and their control requires severe administrative, financial, legal and scientific decisions ○ Maintain health of our planet → uncontrolled deforestation and pollution damage the environment Measurement (through counting) is a legitimate and necessary activity in the study (and management) of society 3 ways to measure variables ○ Measurement of amount ○ Measurement by classification into categories ○ Measurement by ranking Biome: area classified according to the species that live in that location. ○ Temperature range, soil type, and the amount of light and water are unique to a particular place and form the niches for specific species allowing scientists to define the biome. Paper 7 (body temperature) Big data Central statistical concept: variability ○ Successive observations of a system or phenomenon do not produce exactly the same result Aim: Body temperature is a crucial factor used in the diagnosis of acute pathological states, yet the current method for evaluating body temperature is not as precise as it can be. ○ Most studies of body temperature in humans date from 1950 or before and have important limitations ○ Temperatures were measured at varying times of day and in different seasons, using unspecified instruments ○ Sample sizes were small – often in the 10s to 100s of patients Measure temperature: wunderlich’s thermometer Process: ○ patients to be included in the cohort were selected ○ data on each of these patient’s outpatient visits were obtained ○ focused on routine visits during which temperature was expected to be within normal limits ○ Excluded: emergency department visits, during which acute physiological disturbances may affect measured temperatures patients aged less than 18 years and those seen on weekends or outside business hours (7 am-6 pm), to avoid selecting patients seen for non-routine problems implausible recorded temperatures (45°C; 0.04% of the total sample) those visiting clinics for infections visits with ICD-9 (international classification of diseases, ninth revision) codes for infectious diseases, and visits with antibiotics prescribed in the week after the visit  Results: ○ ○ temperature differs according to the site of body site of measurement – mouth, rectum and armpit will give different readings ○ Normal body temperature (NBT) declined with age. ○ African-American women had the highest NBT (NBT higher than white men). Paper 8 (cortisol levels) Animal visitation programme (AVP) provide the general student population an opportunity to engage in hands-on petting of animals for 5 to 45 minutes in small group settings→ It's an approach to stress prevention ○ Ad hoc, e.g., university staff bringing their personal pets to campus ○ Formal, e.g., providing interaction with specially trained therapy dog-handler teams Over the past decade, university students have reported increasingly high levels of academic stress, depressive symptomology, anxiety, and suicidal ideation ○ Effect: lower GPAs, drop out of college Cortisol: steroid hormone that is produced by your 2 adrenal glands, which sit on top of each kidney. ○ When you are stressed, increased cortisol is released into your bloodstream. Randomised controlled trial ○ Randomization reduces bias and provides a rigorous tool to examine cause-effect relationships between an intervention and outcome ○ Researchers must carefully select: Population Interventions to be compared Outcomes of interest Four conditions (10 minutes each) ○ hands-on AVP Research assistants timed entry and exit Four or five students engaged with dogs under supervision of a handler, or cats individually ○ AVP observation This condition studied the experience of waiting in line for their turn in the AVP. Participants were told that they could observe others engage in human animal interaction. But they were to refrain from engaging in physical interaction with animals and remain behind a roped barrier. ○ AVP slideshow Participants were escorted to a room where they viewed a 10-minute slide presentation containing pictures of cats and dogs while indicating their ratings on a checklist. They were also instructed to refrain from interacting with other participants. ○ AVP waitlist Participants were led to a waiting room and asked to store cell phones and reading materials and instructed to refrain from verbal interaction with others. They were reassured that they would experience animal interaction once their data had been collected Recruitment took place the week before final exams, a few days before program implementation, coinciding with a university-wide announcement of the AVP program date and location Data collection ○ Wakeup Cortisol (taken on the day of the intervention) Participants collect a salivary cortisol sample on their own immediately after waking  During the recruitment meeting, participants had already observed a demonstration, practised taking a sample and received verbal instructions They were also given take-home sampling kits They were instructed to store the completed sampling kit in the refrigerator until check-in. ○ Pretest and Posttest Cortisol Participants provided a sample of saliva 15 and 25 minutes after the completion of their 10-minute condition Sampling took place at a designated sampling area where participants also completed a checklist to document behaviour that day pertinent to sampling (e.g., use of medication, food and beverage intake, caffein, nicotine and alcohol intake, exercise, and interactions with animals before programme participation) Results ○ Given that posttest cortisol levels were lowest in the hands-on group and second lowest in the observation group, it is possible that a mechanism by which hands-on petting lowers students’ cortisol levels is through increases in oxytocin. ○ However, perceptual and psychological mechanisms might be involved in the stress-relieving effects of the programme Paper 9 (chloral singing) Over 600 choral singers drawn from English choirs completed the WHOQOL-BREF questionnaire to measure physical, psychological, social and environmental wellbeing, and a twelve-item wellbeing and choral singing scale. ○ The WHOQOL is a quality of life assessment developed by the WHOQOL Group with fifteen international field centres, simultaneously, in an attempt to develop a quality of life assessment that would be applicable cross-culturally They also provided accounts of the effects of choral singing on quality of life, wellbeing and physical health in response to open questions. Statistical technique ○ Principal Component Analysis (PCA) It is a dimensionality reduction and machine learning method used to simplify a large data set into a smaller set while still maintaining significant patterns and trends. It is a machine learning (specifically, unsupervised learning) technique Significant challenge disclosed 1. Affected by enduring mental health problem 2. Affected by significant family/relationship problems 3. Affected by significant physical health issues/disability 4. Recently bereaved Six generative mechanisms (choral singing) 1) Engenders happiness and raised spirits which counteract feelings of sadness and depression 2) Involves focused concentration which blocks preoccupation with sources of worry 3) Involves deep controlled breathing which counteracts anxiety 4) Offers social support and friendship, which ameliorate feelings of isolation and loneliness 5) Involves education and learning which keeps the mind active and counteracts decline of cognitive functions 6) Involves a regular commitment to attend rehearsal which motivates people to avoid being physically inactive Paper 10 (education) Scholarly culture ○ the way of life in homes where books are numerous, esteemed, read and enjoyed ○ provides skills and knowledge that are central to literacy and numeracy ○ There is some evidence that parents’ scholarly culture enhances their children’s educational attainment Parents with high cultural aspirations for their children might well take them to opera, ballet, and museums; buy art for the home; talk to them about literature; listen to classical music with them; keep a copy of Shakespeare and a volume of poetry on the coffee-table; and the like  ○ A home in which books are an integral part of the way of life will encourage children to read for pleasure, thereby providing them with information, vocabulary, imaginative richness, and wide horizons Measure parents’ scholarly culture ○ by the number of books in the home and estimating its effect on children’s education in 27 nations “About how many books were there around in your family’s house when you were 14 years old” ○ It is reported reliably, and has a high test-retest reliability (high consistency of reportage) ○ It is a valid indicator of scholarly culture in the home, and is correlated with other aspects of scholarly culture ○ It has a strong predictive validity as an indicator of parents’ willingness to teach their children ○ It is reliably reported by respondents decades after they left their parents’ homes Results ○ Children from homes with more books had more advantage in years of formal education ○ Higher parent’s education more books in house ○ Advantage of having books at home on children decreases as parents’ education increases Conclusion ○ Books matter hugely for children from low-status homes but they often cannot afford ○ Quality of books available in the home library is a factor that must be considered Big data paper (crime rates) Study problem of crime rate inference of communities Select chicago→ has more homicides and non-negligent manslaughter rates and experienced no decline in past decade Segmented population of chicago according to crime rate Traditional methods of crime rate estimation → Use of demographic information (e.g., population poverty level, socioeconomic disadvantage, racial composition of population) → obtained through a census conducted once every ten years ○ such information does not reflect changes in the community ○ even if two communities are distant, they could be strongly correlated if many people frequently travel between these two communities Data used in the study ○ Data that reflects a city’s dynamics, e.g., traffic flow, human mobility, social media, and crowd-generated Points-of-Interest (POI) → this study uses two – POI and taxi flow (urban data) Taxi is preferred transport by criminals as they offer more privacy and more flexible pick-up and drop-off points POI: Data on GPS coordinates, category (food, shop, transit, education), popularity, reviews. Results ○ high crime in “professional” category ○ low crime in “nightlife” category ○ Overall crime rate positively correlated with taxi flow Big data paper (genomic prediction) Predicting which IVF embryos in a laboratory dish would be most likely to develop Type 1 diabetes, cystic fibrosis, or some other complex diseases The statistical scorecard (or report card) that they produce would enable parents (with the help of their doctors) to avoid embryos with failing grades Predicting other traits are also possible, e.g., height (dwarfism), adult weight, skin tone, personality, intelligence, mathematical ability, etc Statistical forecast ○ For adults, this information is a source of health advice ○ For embryos, it will decide who will be born, and who will be discarded (genetic selection) ○ “selecting embryos for intelligence could boost the resulting child’s IQ by 15 points” ○ “Height is currently the easiest trait to predict.” (34 cm) ○ SF is possible thanks to a growing flood of genetic data collected from large population studies. As statistical models known as predictors gobble up DNA and health information about hundreds of  thousands of people, they’re getting more accurate at spotting the genetic patterns that foreshadow disease risk Big Data paper (Facebook) Facebook investigated the pattern of postings by couples, and revealed that using data from the posts people make, it is able to tell when they fall in love. Background: Couples are formed, and some couples share this news with the world on Facebook by changing statuses from “Single” to “In a relationship”. Facebook explored interactions between couples before and after the relationship begins (denoted as “Day 0”) ○ Post begin to decrease after relationship started → spend more time together physically and neglect online interactions ○ Content of interaction gets sweeter and more positive as relationship progress → count the word expression positive emotions and negative ones FINALS READINGS Personality and Creativity by Angel Aguilar-Alonso investigates the relationship between personality traits and creative behaviour explores how different dimensions of personality—such as psychoticism, extraversion, and neuroticism—affect creative tasks among university students, artists, and other professionals The study involved 400 adult volunteers of various ages, educational levels, and professions (university students, artists, and other professionals) Experimental procedure: ○ The participants were divided into three groups based on their background: university students, artists (writers, musicians, painters, sculptors, architects), and individuals from other professions. ○ Several variables were assessed, including demographics, creativity measures, cognitive abilities, and personality traits. ○ Personality traits such as psychoticism, extraversion, neuroticism, and sincerity were measured using the Eysenck Personality Questionnaire (EPQ). ○ Creativity was assessed through various tasks: Verbal Creativity: Participants designed a crossword puzzle, with points awarded based on letter functions. Creative Originality: Participants completed four drawings, which were evaluated for originality, distinctiveness, and detail. Creative Fluency and Flexibility: Points were given for each distinct feature in the drawings and for flexibility in using different features. ○ Cognitive measures included tests for verbal reasoning, perceptual aptitude, and attention discrimination. ○ Observational conditions were varied across three experimental instructions: 1. Creative Focus: Participants were explicitly instructed to be as creative as possible. 2. Control: They were told to complete personality tests without a focus on creativity. 3. Divided Attention: Participants were asked to be creative while simultaneously counting out loud, introducing a distraction. Data Analysis ○ Factor analysis and multivariate analysis of variance (MANOVA) were applied to explore the relationships among personality traits, creativity scores, and experimental conditions. ○ ANOVA Key findings: 1. Personality and Creativity Links: Personality traits like extraversion and psychoticism showed distinct associations with creativity. Extraverts, for instance, were more fluent in their creative output, while individuals scoring high in psychoticism exhibited both high creative flexibility and detail-oriented perceptions. 2. Influence of Cognitive Abilities: Intelligence (measured by verbal reasoning) correlated with verbal creativity but was less related to drawing-based creative tasks, suggesting that different types of creativity rely on separate cognitive skills. 3. Effect of Task Instructions: Instructions impacted creative fluency; participants tasked with multitasking showed improved fluency in creative tasks, indicating that divided attention can boost certain aspects of creativity 4. Group Comparisons: Students demonstrated higher verbal creativity, while artists were more original and flexible in drawing tasks, suggesting profession-specific creativity patterns. Private Traits and Attributes by Michal Kosinski, David Stillwell, and Thore Graepel explores the predictive power of Facebook "Likes" in determining personal attributes study used a dataset of over 58,000 Facebook users who shared their Likes, demographic information, and psychometric test results How experiment conducted: ○ Data was collected from volunteers who provided their Facebook Likes, along with their demographic profiles and results from psychometric tests such as personality traits (measured by the Big Five model), intelligence, life satisfaction, and substance use. ○ For some variables (e.g., ethnicity), additional data was inferred based on profile pictures or specific Likes. ○ Researchers represented the data in a “user–Like matrix,” with a binary value indicating whether a user liked a particular page. ○ Using singular value decomposition (SVD), they reduced the matrix's dimensionality for improved data processing. ○ They applied logistic and linear regression models to predict specific attributes based on the Likes, using cross-validation for accuracy. Dichotomous (binary) outcomes like gender were predicted with logistic regression, while numeric variables like age were predicted with linear regression. Results ○ The study found high predictive accuracy for attributes such as gender (93%), ethnicity (95%), and political affiliation (85% for Democrat vs. Republican). ○ Personality traits like openness and extraversion were moderately predicted, with correlations close to typical test-retest reliability of personality questionnaires. ○ Even attributes like intelligence and relationship status were inferred accurately, though more complex variables like life satisfaction were less precisely predicted. The study demonstrated that personal attributes could be inferred from relatively simple digital behaviours, highlighting potential privacy risks as individuals might not be aware that such information could be deduced. while digital records (such as Likes) can improve personalised online services, they also raise privacy concerns, as sensitive information about users can be inferred without explicit disclosure Community Gardening by Koay and Dillon explored the mental health benefits of community gardening, focusing on subjective well-being, stress, and resilience among residents in Singapore Experimental Design and Data Collection ○ Participants: 111 adults in Singapore were divided into three groups: community gardeners, individual/home gardeners, and non-gardeners who engaged in other outdoor activities. ○ Methodology: Participants completed a cross-sectional survey assessing subjective well-being, stress, resilience, and related factors such as self-esteem, optimism, and openness. Data collection also included demographic information, nature connection levels, and gardening engagement frequency. ○ Measures: The survey included standardised scales like the Brief Resilience Scale (BRS), Perceived Stress Scale (PSS-10), and Personal Well-Being Index (PWI-A). A single-item scale, Inclusion of Nature in Self (INS), assessed participants' connection to nature. Results and Data Analysis ○ Multivariate Analysis: A MANCOVA was used, controlling for age and nature connection, to analyse group differences in mental health variables. Well-Being: Community gardeners reported significantly higher subjective well-being scores compared to both individual/home gardeners and the non-gardening control group. Resilience: Community gardeners had higher resilience and optimism scores than the control group. Stress: No significant differences in perceived stress were found across the groups. ○ Pearson correlations: Resilience positively correlated with well-being and negatively with stress, supporting the hypothesis that resilience is associated with reduced stress and increased well-being. Key Findings ○ Community gardening was associated with higher subjective well-being, resilience, and optimism, suggesting it may offer unique benefits compared to other forms of gardening or outdoor activities. This supports the idea that community gardens can foster mental health and resilience through social and environmental connections. Personality Traits and Facebook Use explored how Big Five personality traits (Extraversion, Neuroticism, Conscientiousness, Agreeableness, and Openness) and other psychological variables (e.g., self-esteem, narcissism, loneliness, shyness, boredom proneness) influence Facebook usage also investigated how interactions between traits (e.g., Extraversion and Neuroticism) predict behaviour and included both university students and non-students. Participants: 374 participants (190 university students and 184 non-students) aged 18–62 years (Mean age = 26.59). Measures: Big Five traits using the Australian Personality Inventory. Narcissism (Narcissism Personality Inventory-16). Loneliness (UCLA Loneliness Scale). Shyness (Revised Cheek and Buss Shyness Scale). Facebook usage (time spent, daily logins, number of friends). Procedure: Data were collected through online surveys advertised on noticeboards, social media, and emails. Ethical approval was obtained, and participation was voluntary. Key Findings: 1. General Usage Patterns: Students spent more time on Facebook (92.28 mins/day) compared to non-students (71.17 mins/day). Narcissism was the strongest predictor of time spent on Facebook for both groups. Students had more Facebook friends (mean = 437) than non-students (mean = 349). 2. Personality Influences: Extraversion: Predicted the number of Facebook friends across groups. Narcissism: Strong predictor of daily logins for non-students. Agreeableness: Strongest predictor of logins for students.  Interaction of Extraversion and Neuroticism: Predicted Facebook friends for students, suggesting impulsive and complicated personality types were more active in expanding their friend networks. Interaction of Extraversion and Conscientiousness: Predicted Facebook friends for non-students. 3. Other Psychological Variables: Narcissists were likely to spend time on Facebook to control their self-image. Shyness and loneliness had weaker effects on usage compared to other traits. Boredom proneness weakly predicted time spent online but had no relationship with other Facebook usage variables. 4. Interaction Effects: High Extraversion coupled with high Neuroticism led to more Facebook friends among students. Low Conscientiousness with high Neuroticism was linked to fewer friends among non-students. Recommendations ○ Future research should: Use cluster analysis to define personality types more effectively. Investigate longitudinal changes in Facebook behaviour. Assess Facebook's impact on academic performance and well-being. Double-edged Stars investigates the paradoxical effects of receiving Michelin stars on the survival of restaurants in New York City, using a mixed-methods approach explores how third-party evaluations, such as Michelin stars, influence restaurants’ operations, value networks, and likelihood of closure. Methodology 1. Data Collection: ○ Panel dataset of 276 elite New York City restaurants from 2000 to 2019. ○ Archival qualitative materials and 23 interviews with restaurant stakeholders. 2. Independent Variable: ○ Whether a restaurant received a Michelin star 3. Dependent Variable: ○ Restaurant exit (binary: 1 = closed, 0 = active). 4. Analysis Techniques: ○ Parametric proportional hazards regression to estimate effects on restaurant survival. ○ Matched sample analysis and robustness checks to validate results. Key Findings 1. Quantitative Results: ○ Michelin-starred restaurants had a higher likelihood of closing compared to their non-starred counterparts. ○ Parametric analyses indicated that Michelin stars correspond to increased exit risks, consistent across different models. ○ Matched samples reinforced these findings. 2. Qualitative Insights: ○ Upstream Challenges: Increased bargaining power of landlords, suppliers, and employees due to the perceived value of the Michelin star. Higher costs to maintain quality standards and retain key staff. ○ Downstream Challenges: Heightened consumer expectations, leading to operational changes and financial strain. Shifts in customer demographics and preferences, sometimes misaligned with the restaurant’s offerings. 3. Survival Statistics: ○ Of the 276 restaurants analysed, 77 closed during the study period. ○ Michelin-starred restaurants were at greater risk, with 42 of 92 (46%) closing.  Conclusions Paradoxical Effect: While Michelin stars enhance reputation and customer attention, they also impose operational and financial burdens that increase the risk of closure. Value Network Reactivity: Changes in upstream (suppliers, landlords) and downstream (consumers) relationships significantly impact restaurants’ ability to sustain operations. Managerial Implications: Businesses must carefully consider the trade-offs of pursuing prestigious evaluations. Effectiveness of Online Marketing Tools ‘customers journey’ portray the ‘tour’ a customer goes through to purchase or order a product/service of a particular firm → reveals what sorts of contact points are responsible for a prospective consumer to reach the website of the company (e.g advertisement, online marketing measures like email marketing, chat section, word of mouth, radio, banners, layer ads, Internet or websites of others firms) We analysed revising practices at Amazon.com and bn.com and discovered that customer appraisals tend to be constructive at both sites and that they are more comprehensive at Amazon.com. This confirmation suggests that customer endorsement touches consumer purchasing performance at two internet retail sites. Problem statement ○ explores the impact of different online marketing strategies on consumer decision-making ○ H1: Role of email advertisement in influencing the decision of prospective buyers. H2: Role of online chat section in influencing the decision of prospective buyers. H3: Role of word of mouth in influencing the decision of prospective buyers. Data Analysis ○ Data was collected through a survey conducted in December 2019 using a questionnaire distributed to 100 participants in Gurugram, Haryana. ○ Out of these, 95 responses were fully completed. The questionnaire used was close-ended with multiple choices on a 5-point Likert scale to gauge responses, and one-sample t-tests were performed to evaluate the effectiveness of each tool. Experiment and Methodology ○ Participants, selected via random sampling, completed a survey aimed at identifying the impact of each marketing tool on purchasing decisions. ○ The t-tests compared each tool's mean response to a test value of 4 (strong agreement) to determine the influence level. Results: The analysis showed that: One sample t-test: H1: rejected, H2: rejected, H3 accepted Word of mouth was the most effective, with higher mean scores and significant positive t-test results, indicating strong influence on consumer decisions. Email advertisements and online chat sections had lower means and significant negative t-test results, suggesting a lesser impact. Further research ○ public’s access to wireless Internet in urban and rural areas ○ Another possible area of research for online marketing can be Facebook, Twitter, etc., Mental Illness-Stigmatizing Beliefs assess beliefs about mental illness in a multiethnic Singaporean population, focusing on whether people view mental illness symptoms on a continuum and how these views relate to stigma Experiment method: ○ This community-based cross-sectional survey included 3,006 Singaporean residents aged 18-65, selected using stratified random sampling to ensure representation across ethnic and age groups. ○ Participants were presented with case vignettes depicting five mental disorders (alcohol abuse, dementia, depression, schizophrenia, and OCD), matched to the respondent’s gender and ethnicity for cultural relatability. ○ Responses were collected through structured interviews covering continuum beliefs, personal stigma, social distance, and causal beliefs. Data Analysis: ○ Descriptive statistics and multiple linear regression were used to identify demographic and belief-based correlates of continuum views.  Mean Scores and Standard Deviations: These were used to capture central tendencies and variations in stigma dimensions, such as social distance, "weak-not-sick" beliefs, and "dangerous/unpredictable" perceptions. For instance: Social Distance scale and continuum belief scale ○ Regression analysis examined factors like employment, previous contact with mental illness, and beliefs about causes. Results: ○ Findings showed that continuum beliefs were more common for depression (57.9%) and dementia (46.8%) than for other disorders. ○ Those with prior contact with mental illness, students, and unemployed individuals more strongly endorsed continuum views. ○ Continuum beliefs were linked to lower social distancing in schizophrenia, OCD, and alcohol abuse but correlated with greater perception of “weakness” for dementia and schizophrenia