Undergraduate Research · Consumer Behavior · SPSS + AMOS

Tracing the Novelty,
Savoring the Future

A survey of 1,030 Nanjing residents on willingness to buy novelty-flavored beverages — modeled with the Theory of Planned Behavior, K-means clustering and SEM.

0valid responses
0%valid rate
α=0Cronbach's alpha
0KMO measure
Paper cover — cartoon novelty drinks
Crafted by Zhijie Wu·zhijiewu1105@gmail.com
Word cloud of novelty beverage keywords from the paper
Market Context · Novelty Beverages · Nanjing
01 · Research Background

When weird flavors go viral

From salted-egg-yolk soda to cilantro juice, novelty flavors have become the beverage industry's most talked-about frontier — but virality does not guarantee repeat purchase.

¥21.61 bn

Beverage GMV on major e-commerce platforms, Jan–Jul 2024 +27.8% YoY

+238.36%

Quarterly sales growth of niche flavors (cilantro, houttuynia, luosifen, durian) on Taobao, Q1 2024

8,000 + orders

Online orders of Huiyuan 100% cilantro juice within its first week — despite polarized reviews

“Why do consumers pay for novelty-flavored beverages — and how does the decision logic systematically differ across consumer groups?”
Core research question · Nanjing, China
02 · Theoretical Framework

The TPB engine, made clickable

The questionnaire operationalizes the Theory of Planned Behavior across four constructs, five Likert items each. Select a node to inspect its definition, sample items and evidence.

β = 0.368 β = 0.355 β = 0.273 only 34% convert Attitude Subjective Norm Perceived Control Intention Behavior
03 · Methodology

From questionnaire to model

A multi-stage sampling design across six Nanjing districts, followed by a full statistical pipeline in SPSS and AMOS.

STEP 01

Instrument design

TPB-based questionnaire; 4 constructs × 5 Likert items, refined through an 87-response pilot.

Attitude scale results — radial chart from the paper
STEP 02

Multi-stage sampling

Stratified + Lahiri + SRS; six districts — Lishui, Jiangning, Qinhuai, Jianye, Qixia, Luhe.

Nanjing administrative map
STEP 03

Fieldwork

Online + offline interception surveys in public spaces; dual data entry and four validity screens.

STEP 04

Analysis

SPSS + AMOS: reliability & validity, EFA, ordered-logistic regression, K-means, MANOVA, SEM.

K-means cluster scatter plot from the paper
0distributed
0valid
0%valid rate
Cronbach's α = 0.885 KMO = 0.981 Pilot n = 87 6 districts · 9.55M population frame
α ≥ 0.8 and KMO ≥ 0.8 mean the questionnaire is reliable and the data suit factor analysis.
04 · Data Insights

What 1,030 residents told us

Social influence runs high, purchase conditions are favorable — yet conversion lags far behind intention.

0%of intention converts to action
Can't find a store27%
Taste below expectation19%
Promo buyers → loyal22%
The three frictions behind the gap: limited access, unmet expectations and promo-dependence.

Subjective norm — % agreement

Share rating 4–5 on 5-point items
All seven items above 72% — social pressure is strong across the board.

Perceived behavioral control — means

5-point scale, n = 1,030
All means sit at 3.84–3.96 of 5 — buying feels easy, affordable and within reach.

Purchase intention — means

5-point scale, n = 1,030
Means near 4 of 5 — stated willingness runs high in every form.

Monthly disposable income

Sample composition, %
65% of the sample earns ¥3,000–8,000/month — the market's core spending group.
05 · Consumer Segments

Three tribes of novelty drinkers

K-means clustering (elbow method, silhouette 0.389) separates consumers into three sharply distinct groups.

40.78%
Cluster 1 · Engaged Explorers

The social-driven core

  • Mid income & education; high liking and trial intent
  • Highly swayed by friends, family and social media
  • The core growth audience for novelty launches
47.33%
Cluster 2 · Premium Quality-Seekers

The quality-led potential

  • Higher education & income; uniqueness and brand image matter
  • Decisions still driven by family and social circles
  • Natural target for premium, design-led lines
11.89%
Cluster 3 · Disengaged Observers

The hesitant minority

  • Younger, lower income; interest and intent far below others
  • Low scores on attitude, norm and control alike
  • Not a priority segment for current campaigns

Segment share of sample

K-means, k = 3 · n = 1,030
88% of consumers fall into the two high-potential groups.

Attitude & trial intent by segment

Cluster centroids, 5-point scale
Cluster 3 scores far lower on every attitude item — a clear engagement divide.
06 · Structural Equation Model

The model holds — and explains

All six hypothesized paths are significant and positive. Attitude is the strongest direct driver of intention, followed by subjective norm and perceived control.

0.368 *** 0.355 *** 0.273 *** Attitude 6 observed items Subjective Norm 7 observed items Perceived Control 7 observed items Purchase Intention 7 observed items Behavior actual purchase intention–action gap
Numbers on the arrows are standardized path coefficients (β) — larger means stronger influence; *** = p < 0.01. Chips below show the model's fit indices with their accepted thresholds.
0.986χ²/df (<3)
0.948GFI (>0.9)
≈0.000RMSEA (<0.10)
1.001CFI (>0.9)
0.948NFI (>0.9)
1.001NNFI (>0.9)
07 · Findings & Implications

Four takeaways, one playbook

Novelty beverages are bought socially, judged on experience, and lost at the last mile.

1

Subjective norms dominate

0%
refer to others' opinions before buying
KOL influence 41% among youth
2

Novelty & convenience convert

18% → 27%
monthly repurchase rate with 30-minute delivery
3

The intention–behavior gap

0%
of strong intention converts to real purchase
27% can't find a store
4

Health & function emerge

0%
spontaneously ask for healthier recipes
38% young women eye beauty drinks

For brands

Segmented operationsPremium collabs · scenario lines · trial packs
Social-first growthReferral rewards · campus communities · UGC
Omnichannel convenienceSmart vending · 30-min delivery · mini-program
Continuous noveltyFunctional ingredients · mystery flavors

For consumers

Check ingredients & certifications Start with small packs Brand over flashy packaging Convenience & fair price
08 · Appendix

Survey instruments & paper artifacts

Read the full paperOriginal report (Chinese) · PDF · all tables, models & interview records included Download PDF

Novelty-Flavored Beverage Consumption Survey

Final instrument · anonymous · 5-point Likert matrices (1 = strongly disagree, 5 = strongly agree) · 猎奇口味饮料消费倾向调查问卷
Q1 Gender 性别 · Male 男 / Female 女 Q2 Age 年龄 · 18以下 / 18–25 / 26–30 / 31–40 / 41–50 / 51–60 / 60+ Q3 Education 学历 · 初中及以下 → 研究生及以上 Q4 Occupation 职业 · 22 categories Q5 City 城市 (fill-in)

Section A · Attitude toward novelty beverages 行为态度

Item 题项12345
High overall liking for novelty beverages我对猎奇口味饮料的整体喜好程度很高
The taste is truly unique猎奇口味饮料的口味非常独特
More appealing than regular beverages猎奇口味饮料比普通饮料更有吸引力
Satisfies my curiosity猎奇口味饮料能够满足我的好奇心
Very willing to try new novelty flavors我非常愿意尝试新的猎奇口味饮料
The packaging design is creative猎奇口味饮料的包装设计很有创意
The brand image attracts me猎奇口味饮料的品牌形象很吸引我

Section B · Subjective norm 主观规范

Item 题项12345
My family strongly supports my purchase我的家人非常支持我购买猎奇口味饮料
Friends often recommend them to me我的朋友经常推荐我购买猎奇口味饮料
Others' opinions change my purchase decisions我会因为周围人的意见而改变购买决策
I feel everyone around me is buying them我觉得周围人都在购买猎奇口味饮料
Buying them is a trendy behavior购买猎奇口味饮料是一种潮流行为
My colleagues rate them highly我的同事对猎奇口味饮料的评价很高
I refer to social media recommendations我会参考社交媒体上的相关推荐

Section C · Perceived behavioral control 感知行为控制

Item 题项12345
Buying them is very convenient购买猎奇口味饮料非常方便
The price is reasonable猎奇口味饮料的价格非常合理
I encounter almost no barriers购买时几乎不会遇到障碍
Sales channels are abundant猎奇口味饮料的销售渠道非常多
Packaging strongly influences my decision包装设计对我的购买决策影响非常大
Easy to find in supermarkets我很容易在超市找到猎奇口味饮料
Promotions attract me猎奇口味饮料的促销活动很吸引我

Section D · Purchase intention 购买意图

Item 题项12345
I will definitely buy in the future我未来一定会购买猎奇口味饮料
Very willing to recommend to others我非常愿意向他人推荐猎奇口味饮料
Plan to buy within a month我计划在未来一个月内购买猎奇口味饮料
Buying them is a fun experience购买猎奇口味饮料是一种非常有趣的体验
A product well worth trying猎奇口味饮料是一种非常值得尝试的产品
I regularly follow new launches我会定期关注猎奇口味饮料的新品信息
Promotions increase my purchase volume我会因为优惠活动而增加购买量

Open questions 开放题

Q11 — Suggestions for improving novelty beverages 您对猎奇口味饮料的改进建议?
Q12 — Suggestions for the market's development 您对猎奇口味饮料市场的发展建议?

Research framework — original figure from the paper

研究框架图 · from problem statement → design & fieldwork → analysis → conclusions & outlook
Research framework diagram from the paper

TPB concept diagram — original figure from the paper

TPB 理论概念图 · beliefs → attitude / subjective norm / perceived control → intention → behavior
TPB concept diagram from the paper

Semi-structured interview outline

深度访谈提纲 · conducted with brand R&D staff, food bloggers and novelty-drink triers
1
How often do you usually drink beverages?您平时喝饮料的频率大概是怎样的?
2
Have you tried novelty flavors — e.g. insect or herbal-bitter ones?您是否尝试过猎奇口味的饮料,比如昆虫风味、草本苦味等?
3
Through which channels do you learn about them — social media, friends, offline ads?您通常通过哪些渠道了解到猎奇口味饮料?
4
What attracts you to try them — novelty, brand marketing, others' recommendations?吸引您尝试猎奇口味饮料的原因是什么?
5
When buying, do you value brand, price, or the flavor itself more?您购买时更注重品牌、价格,还是口味本身?
6
In which occasions do you choose them — gatherings, leisure, work?您一般在什么场景下会选择喝猎奇口味的饮料?
7
Are you satisfied with the current variety? What flavors do you wish to see?您对目前市场上的种类和丰富度满意吗,希望增加哪些口味?
8
Would packaging design spark your interest in a product?您是否会因为包装设计而对一款猎奇口味饮料产生兴趣?
9
What price range feels reasonable to you?价格在什么范围内您觉得比较合理?
10
Would you recommend them to people around you, and why?您会向身边人推荐猎奇口味的饮料吗,基于什么原因?