Transcript My presentation - University of Pittsburgh

An Experimental Study of Antecedents
and Consequences of Pop-Up Ad
Intrusiveness: A Study in Progress
Dennis Galletta, University of Pittsburgh
Scott McCoy, College of William & Mary
Andrea Everard, University of Delaware
Peter Polak, University of Miami
(author order to be determined)
Syracuse University Seminar
October 24, 2005
Slide file # 34-17
Page 34-17
Agenda
 Problem Statement
 Research Questions
 Background and
Hypotheses
Slide file # 34-17
 Research Methodology
 Data Collection
 Preliminary Results
 Future Steps and Studies
Page 34-17
Online Ads: The Good News

Internet advertising reached $5.8 B in 1st half of 2005
(PWC, 2005)
– Annualized, that is $3 B over the peak of the dot-com boom
– Increased 26% in 2005, coming from 40% increase in 2004
– The picture is bright indeed!


Several formats: e.g.:
in-line, pop-up, and pop-under
Early research
– Expected ads to be less intrusive (no forced watching as on
TV) (Rust & Varki, 1996)
– effective in creating brand awareness and positive attitudes
(Briggs & Hollis, 1997)
Slide file # 34-17
Page 34-17
Online Ads: The Bad News

Studies report that consumers despise
these intrusive and annoying
advertisements
Ads are described as disturbing (Reed, 1999),
nonsensical, uninformative, forgettable,
ineffective
Terms such as “violated” and “molested” are
used by consumers to describe their
experiences with advertisements
(Reed, 1999; Rettie, 2001; Li et al. 2002; Chan et al. 2004)

(Bulik, 2000)

–
Slide file # 34-17
Source: Wegert (2002)
Page 34-17
Blame Spreads Around

Slide file # 34-17
Even though Google has no pop-ups, “spyware” leads people to blame them.
Page 34-17
E-Commerce Background
Literature



Bauer & Greyser (1968) recognized intrusiveness as
a major cause of advertising annoyance.
Rust & Varki (1996) expected online ads to be less
intrusive due to greater user control
Li, Edwards, and Lee (2002) found that
– online consumers are goal-oriented and perceive ads to be
even more intrusive than those in other media.
– Consumers develop negative attitudes towards the ads
and/or avoid them altogether.
– These negative attitudes can lead to intentions not to
revisit a site with ads
Slide file # 34-17
Page 34-17
Advice From Psychology I



Theories of Attention (Broadbent 1958; Treisman 1988)
People monitor information through
multiple channels, but only attend to one
People remember less when interrupted
– Level of processing hypothesis (Craik, 1982)
– Reduced processing time hypothesis (Craik et al.,
1996)
– Fragmentation hypothesis
Slide file # 34-17
(Naveh-Benjamin 2000)
Page 34-17
Advice from Psychology II

Dual-Process theory of person
perception
A person:
(e.g., Smith & DeCoster, 2000; Hastie & Kumar, 1979)
– recognizes and internalizes general
regularities with one process, but
– Requires a second, more conscious and
effortful process for unique or novel
events (such as non-congruent info)
Slide file # 34-17
Page 34-17
Bridging to The
Affective Domain

Multiple non-automatic tasks require:
(Kahneman, 1973; Kahneman & Treisman, 1984)
– Divided attention
– Active filtration of irrelevant stimuli



Slide file # 34-17
Different channels interrupt each other very
little (Wickens, 1980)
Repeated same-channel interruptions force
much effort, which is undesirable to an
organism (Reid & Nygren, 1988)
Greater demands lead to negative affect
(Eggemeier et al., 1991)
Page 34-17
Previous Study





Slide file # 34-17
We had a large number of variables
Too many hypotheses, too complex
Findings were hard to internalize, hard
to organize
We were rejected from some of the
finest journals
For CACM, (forthcoming) we simplified
to three categories of questions
Page 34-17
Previous Study-Questions
1.
2.
3.
Slide file # 34-17
How intrusive are online ads? Will the ads
lead users to not return to the site hosting
the ads?
Are some forms of online ads more
intrusive than others? Will it make a
difference if the ads are not related to the
subject matter of the site?
Will online ads interfere with users’ ability
to remember site content? Which types will
provide most interference?
Page 34-17
Research Methodology







Slide file # 34-17
Laboratory Experiment conducted with
undergraduates
Mexico (n=208) and US (n=209)
210 Females, 207 Males
Fictitious website created and accessed via
server in US
Subjects randomly assigned to treatment
groups
Subjects completed 9 search tasks
Ads were displayed during search tasks
Page 34-17
Methodology

Dependent Variables
– Behavioral Intentions
– Retention of Site Content
– Retention of Ad Content
– Perceived Intrusiveness of Ad
Slide file # 34-17
Page 34-17
Design

Factors
– 4 ad categories (controlling for size,
location)
pop-up
 Pop-under (exploratory treatment)
 In-line
 None

– Congruent or not
– Animated or not
Slide file # 34-17
Page 34-17
1. Are Ads Intrusive?




Ads take up space on our screens
(Reed, 1999; Rettie, 2001;
Li et al. 2002; Chan et al. 2004; Bulik, 2000; Wegert, 2002)
Goal orientation (Li et al., 2002; Rapoza, 2004)
Additional effort from competing stimuli
require more effort (Kahneman & Treisman 1984, Wickens 1980)
Our results:
– Measures of intrusiveness seemed to be in the
middle of Likert-type scales
– But without ads, subjects were 11% more likely to
return or recommend (p<.01)
– For a site like Amazon, that could amount to 5
million visitors
Slide file # 34-17
Page 34-17
2. Types – Styles




In-line – a graphical element as part of the
site (e.g., banner)
Pop-over – a separate window on top of the
site
Pop-under – a separate window covered by
the site (but it is revealed later)
In the literature, some effects were found –
in-lines were better received than pop-ups
(Denes, 2001)
Slide file # 34-17
Page 34-17
2. Results – Styles
Intrusiveness of Various Forms of Web Advertisements
29
28.0
27
26.1
25
23
21
21.1
19
17
15
In-Line
Slide file # 34-17
Pop-Up
Pop-Under
Pop-ups were 24% more intrusive than in-line ads (p<.001);
Pop-unders were 33% more intrusive (p<.001)
Page 34-17
2. Results – Styles
Intentions to Revisit the Main Site
15
14.5
14.5
14
13.5
13.3
13
12.9
12.7
12.5
12
11.5
In-Line
Pop-Up
Pop-Under
No Ad
Mode NS; Existence is important (p<.01)
Slide file # 34-17
Page 34-17
2. Types – Animation


Animation is used in some ads to
attract attention
Attracting attention is similar to
DIStracting a person from the site
(Hong
et al., 2004; Zhang, 2000)

Slide file # 34-17
However, that distraction does not pay
off in remembering the ad (Hong et al., 2004)
Page 34-17
Static vs Animated
Slide file # 34-17
Page 34-17
2. Types – Congruence



Some ads are congruent with site
content (e.g., film ad on camera site)
Many are not (e.g., life insurance ad
on camera site)
The former is likely to be of interest;
The latter is likely to be intrusive
(Chan, Dodd,
and Stevens, 2004)
Slide file # 34-17
Page 34-17
Congruent vs Non-congruent
Slide file # 34-17
Page 34-17
2. Results – Animation &
Congruence


Slide file # 34-17
Animation did not affect intrusiveness
or irritation
Congruence did not affect
intrusiveness or irritation
Page 34-17
3. User’s Memory of the Site

Cognitive effort is increased by the ads
(Chan, et al., 2004 Kahneman & Treisman 1984, Wickens 1980)

Memory (retention) of the site was
predicted to be adversely impacted by
– Existence of ads
– Congruence (perhaps counter-intuitive)
– Animation

Slide file # 34-17
Also examined: type of ad
Page 34-17
3. Results




Slide file # 34-17
Retention was not affected by the ads
(perhaps users ignore them as stated by
Chan et al., 2004)
Retention was 3.5% higher for noncongruent ads (p<.005)
Retention was not affected by animation
Retention was 3.4% higher for in-line ads
than pop-up ads (p<.022)
Page 34-17
Where Did That Study Get
Us?


Not very far
A new study was needed
– To understand ads more thoroughly
– To understand new factors
– To explore underlying generic issues
Slide file # 34-17
Page 34-17
New Study: Perspectives
Covered and Goals

Site host
– Either a vendor or an information site
– Goal: wants users to return
– Shopping site has further goal of making sales,
but key to profitability is repeat sales—must
return!

Advertiser
– Wants “top of mind” (and clicks & sales)

User wants big payoff with low effort
– Anything else is intrusive
– Make too much trouble? User won’t return
Slide file # 34-17
Page 34-17
Model – First Cut
Intrusiveness
Irritation
Recall of
the Ads
Slide file # 34-17
Attitude
toward
Site
Behavioral
Intentions
Site’s Point of View
Advertiser’s
Point of View
Page 34-17
Approach




Slide file # 34-17
To support the model, we wanted to create
ads with a variety of characteristics
In the CACM study, we varied existence,
animation, and congruence
This time we varied positioning, size, extent to
which the ad obscured site content, and, if
obscuring content, method by which the site’s
content could be seen
We wanted to control nearly everything
Page 34-17
Methodology


Similar to previous study
Created new ad types
– Pop-up ad that obscured content,
needing to be closed
– Equivalent-sized pop-up ad that did not
– Equivalent-sized in-line ad that did not
– Equivalent-sized in-line ad that did, but
disappeared or twirled away
Slide file # 34-17
Page 34-17
Our Data Set


Slide file # 34-17
263 undergraduate business students
at 3 universities in the U.S.
43% were females, 57% males
Page 34-17
Popup
Slide file # 34-17
Page 34-17
In-line
Slide file # 34-17
Page 34-17
Pop-up Obscuring
Slide file # 34-17
Page 34-17
Large Window Pop-up
Slide file # 34-17
Page 34-17
Large Window in-Line
Slide file # 34-17
Page 34-17
Our Observations

Two issues seemed to come out:
– Whether the window obscured site
content
– Whether the ad was under user control
(could be closed by the user)
– We predicted that an ad that obscures
needs more control
Slide file # 34-17
Page 34-17
Revised Model
Obscuring
of Site
Content
Intrusiveness
Lack of
User
Control
Irritation
Attitude
toward
Site
Behavioral
Intentions
Recall of
the Ads
Slide file # 34-17
Page 34-17
Results of Simple
Preliminary Analysis
Obscuring
of Site
Content
Lack of
User
Control
-.339
.789
Intrusiveness
Irritation
.302
.489
Attitude
toward
Site
Behavioral
Intentions
Recall of
the Ads
Interaction of Obscuring
and Lack of Control on
Intrusiveness: F=12.3;
p=.001
Slide file # 34-17
Page 34-17
Interaction
Interaction of Obstruction and
Control
Intrusiveness
6
5
NonObstructing
Obstructing
4
3
2
No Control
Slide file # 34-17
Control
Page 34-17
Further Work Needed



Slide file # 34-17
Analyze model using SEM (after
hurricane season)
Work on justifications of hypotheses
Write paper
Page 34-17
Limitations

Subjects were college students, so results
might not generalize
– But Voich (1995) found students to be
representative of values and beliefs of individuals
in a variety of occupations
– Materials were designed to tap “invariant” (Simon
1990) activities such as aspects of Web use.

Slide file # 34-17
Site was not real
Page 34-17
Future Research


Employ real sites, real ads chosen in
some systematic way
Investigate effects on
– Trust
– Attribution
– Overall website quality
– Purchase intentions or actual purchases
Slide file # 34-17
Page 34-17
Concluding Remarks



Slide file # 34-17
We have an advertising “arms race”
This topic could become obsolete if
advertisers lose
If both escalate the battle, controlled
studies should tell us what we are
getting into
Page 34-17
An Experimental Study of Antecedents
and Consequences of Pop-Up Ad
Intrusiveness: A Study in Progress
Dennis Galletta, University of Pittsburgh
Scott McCoy, College of William & Mary
Andrea Everard, University of Delaware
Peter Polak, University of Miami
(author order to be determined)
Syracuse University Seminar
October 24, 2005
Slide file # 34-17
Page 34-17
Spanish Translations
Slide file # 34-17
Page 34-17