ACM Journal of Data and Information Quality

Providing Research and Tools for Better Data

01 Emma Alexa (.) [ (1.0) play beat the intro
02 Carl [is it called beat the intro?
03 (1.9)
04 Alexa you want to hear a station for b b intro. [(0.4)  right?
05 Emma [ no
06 (1.1)
07 Emma no (.) I don’t Alex(h)uh (0.5) (h)No,
08 (1.3)
09 Alexa alright.

Figure 3. Emma and Carl try to start a game called Beat the Intro.

ACM JDIQ is a multi-
disciplinary journal
that attracts papers
ranging from

(line 07). We see this kind of repetition and variation frequently when users are trying to deal with trouble in use. Alexa then responds with another similar question-categorization, “I don’t have the answer to that question” (line 09).

which emerge frequently to smooth everyday interactions along [ 10]. This emerges more broadly as a kind of “politics of control” that is worked out as part of the life of the home

theoretical research
to algorithmic solutions

Fragment 3. We want to draw
attention to the response from Alexa
on line 04 and what it lets Emma
do next after having instructed
Alexa to play Beat the Intro. The
device’s response here incorporates
a transcription of the result of its
speech-recognition process, “b b
intro.” Although it is actually a
mis transcription by the Echo in this
case, the response nevertheless builds
places to go with her next turn-at-talk,
so she repeats Susan’s instruction with
a slight variation: “Set a family quiz”
[ 3]. Competent conversationalists
routinely perform remedial action to
repair emerging misunderstandings
between themselves and others [ 7].

to empirical research
to experiential
evaluations. Its
mission is to publish

Liam attempts another variation that
displays his recognition of the situation
with its troublesome character and
transforms the attempt at a further
instruction to Alexa into something
amusing: “Please set a family quiz”
But voice-driven devices seem poorly
designed to live in this world of
constant “fixing,” and as a result it is
users of them who are thus seeking to
repair various sense-making problems
that are encountered.

high impact articles

(line 11). This differs from Fragment

contributing to the
field of data and
information quality (IQ).

1 slightly in that here Liam embeds a humorous turn of phrase into something designed for a response from the Echo, the evidence of which is the shared silence of 1.2s on line 13 (compare with the absence of anything similar on lines

09–11 of Fragment 1). Finally, there is
another similar response from Alexa
and another even more pared-down
attempt from Carl: “Alexa, family quiz”
Next we want to contrast Fragment
2 with an alternate way these kinds of
designed requests and responses might
play. While what happens in Fragment
3 (Figure 3) below is also a “failure,”
it turns out quite differently for the
family. The family is trying a different
Alexa skill, a game of Beat the Intro,
which plays just the beginning of a song
and then players must guess the song or
artist name.

(lines 14 and 18).

This is an example of collaborative repair by the family. Furthermore, these collaboratively produced, minutely varied repetitions of the request to “set a family quiz” seem to be closely aligned with the repeated unhelpfulness of the responses from Alexa. As a social environment, home life frequently turns on offers of help (both explicit and implicit),

For further information
or to submit your
manuscript,
visit jdiq.acm.org

We think the concept of conversational design needs to be revised, specifically by talking about sequentially organized moves around request and response.