The effect of line orientation on the recording of time-intensity perception of sweetener solutions

Food QualityandPrcfmtlce6 (1995) 121-126 0 1995 Elsevier Science Limited Printed in Great Britain. All rights reserved 095c-3293/95/$9.50+.00

0950-3293(94)00021-2

ELSEVIER

THE EFFECTOF LINE ORIENTATIONON THE RECORDING OF TIME-INTENSITY PERCEPTIONOF SWEETENER SOLUTIONS L. M. Duizer,aK. Bloomb & C. J. Findlaya “Compusense Inc., 150 Research Lane, Guelph, Ontario, Canada, NlG 4T2 ‘department of Psychology, University of Waterloo, Waterloo, Ontario, Canada, N2L 3Gl

(Accqbted11 October1994)

and different

ABSTRACT

scales. Initially,

A trained time-intensity panel

was used to valuate

indicate

the

pen

effect of scale orientation on time-intensity responses. Equisweet samples of aspartame, aces&fame k, sucralose and

eliminate

the transformation

tical

analysis.

commitment

distraction

to the panelists

Birch

1980;

paper.

& Munton

an instrument

To

of the

(1981)

containing

de-

a taste

moved the dial

of the data into a useable

This

procedure

and increased

With the advent

required

form was

a large

the potential

of personal

198Os, computerized

computers

time-intensity

time

for errors. in the early

software

programs

were developed. These programs were used to collect data in a form which could be easily analysed. Programs have differed responses.

in the type of instrument

For example,

used to record

et al., (1985)

Guinard

intro-

duced the use of a joystick which could be moved within a vertical slot, while Lee (1985) attached a game

INTRODUCTION

paddle

sensory

evaluation

allows for the measurement the

time-intensity

methodological Time-intensity

(Schwartz,

strip chart

cumbersome. The curves were measured manually, and the results were then entered into a computer for statis-

Keywords: time-intensity; sweetness; sweeteners; scale orientation.

recording

vertically

1978;

tensity curves of all responses. Although the automated approaches collected time-intensity data efficiently,

multi-attribute evaluations of taste.

and

a

on moving

of chart paper,

the SMURF,

to

& Pangborn,

‘dial box’ and ‘taste dial’. The panelists

the presentation of the scales within and amongst panelists. These results demonstrate the use of time-intensity scales on two dimensions and suggest the possibility of

texture

or

panelists by moving

clockwise in response to changes in sensory perception, and a concealed strip chart recorder provided time-in-

Intensity responses on the horizontal scale. The parame ters of Decrease Angle, Decrease Area and Area Under the Curve were also significantly larger when vertical scales were used than when horizontal scales were used. We suggest that diff erences can be minimized by anchoring refmence samples to the scales and by counterbalancing

in the

1983)

the possible

movement veloped

required

perceptions

(Larson-Powers

et al., 1984) & Lee,

of time-intensity

procedures

time-intensity

horizontally

Schmitt

horizontal and vertical scales yielded similar results. Howevq Maximum Intensity responses on the vertical scale were approximately 13% greater than Maximum

Time-intensity

automated

their

DuBois

9 % sucrose were presented to 10 panelists fm evaluation on both horizontal and vertical scales. For the most part,

visual representations

is a method

of the temporal

flavour

of food.

responses

have

which

At present,

changes

Devices

With the latter instrument,

there are two commercially

ware packages

for

for the measurement

taste perceptions.

undergone

pan-

Both computer

available

soft-

of time-intensity

packages

are used to

collect data over a programmed period of time. However, they differ in the visual orientation of the scale on which the panelists indicate the intensity of their per-

improvement since their inception. data were originally collected by timing

panelists with a stop watch as they recorded their responses to intensity on graph paper (Neilson, 1957). Since then, researchers have developed both automated and computerized time-intensity These procedures have employed different

to a computer.

elists turned a dial on a ‘game paddle’ to move an ‘x’ along a horizontal line shown on the computer monitor.

ception. The PSA-system provides a vertical scale on which the panelists indicate their perceptions of the attribute (Cliff & Heymann, 1993). In contrast, the

procedures. instruments

CSATpATMtemporal 121

profile analysis module

developed

by

122

L. M. Duizer, K. Bloom, C. J. Findlay

Compusense,

provides

a horizontal

scale (equivalent

to

60 pixels) on which the panelists indicate their perception. Compusense

is currently

simultaneously attributes.

measure

This

requires

scales displayed

horizontal

effect

time-intensity

data

scale with the judgement

a vertical

scale to determine

affects response

a

using

scale orientation

eners

on both

testing

AND

eners

in

isosweet paired

distilled

water.

to 9% sucrose,

sweetener

aspartame,

All

testing.

k; 0.018%

each sweetener

samples.

During

were evaluated

each

with a 5-

The rating scale appeared

as a horizontal

line for four samples and

order.

four samples in a ran-

During

the panelists

crackers

scales. Four

conducted.

samples

the samples,

the 5-min break

cleansed

their palate

and distilled water.

concentrations

0.05%

sucralose.

were poured

were

program of the

of each

Each

aspartame;

Twenty

analysis parameter curve

0.08%

millilitres

of

(CSA

is defined

were

were

curves V4.3).

in Table

horizontally

from

CSATrATM

1 provides and

a

each

1. of the time-intensity

to determine

rated the four sweeteners

the

parameters

parameters

analysed

extracted

using Figure

time-intensity

of the eight

presented

into a 50 ml cup labelled

parameters

time-intensity

diagram

earlier

through

The concentrations

time-intensity

individual

acesulfame

using food grade sweet-

as determined

were: 9% sucrose;

acesulfame

eight

present.

the four sweet-

and vertical

were

and balanced

with unsalted

Eight (sucrose,

were prepared

comparison

line, Testing

Analysis of the time-intensity data

solutions

k and sucralose)

evaluated

the horizontal

min break between

between

METHODS

perception.

was no longer

as a vertical line for the remaining

Sample preparation Four model

the time-intensity

the panelists

of testing

session,

domized

MATERIALS

perception

testing,

on the monitor

to taste perception.

along

in sweetness

of each sample ended when the panelists moved the cursor back to zero on the time-intensity line, indicat-

replications

using

of sweetness

whether

the changes

During

In the present

of sweetness

to move a cursor

ing that sweetness

The

in the literature.

judgements

to

sensory

and vertical

monitor.

on the resultant

reported

study we compared horizontal

both

a system

dependent

on the computer

of scale orientation has not been

developing

two time

mouse

reflecting

differently or vertically.

whether

panelists

when the scale was Orientation

(2) X

with a random Sdigit blinding code. Each covered with a lid and a straw was inserted

cup was through

Sweeteners (4) repeated-measures analyses of variance were conducted for each parameter. When significant

a hole

at room

differences

in the lid. All sweeteners

were served

eners

temperature.

between

occurred,

measures

or amongst

multiple

orientations

comparisons

and sweet-

with repeated

t-tests were conducted.

Time-intensity evaluations All training

and testing

sessions were conducted

Compusense Sensory Research Canada). Ten trained panelists, time-intensity

testing of sweeteners

at the

Centre (Guelph, experienced in participated

in this

RESULTS

AND

When

across

the four sweeteners,

from

vertical

averaged

ratings

differed

study. Prior to testing, the panelists attended eight l-h training sessions to familiarize themselves with the use

eight time-intensity p = 0.003); AUC

of horizontal

ANGLE (F(1,9)

and vertical

line scales.

During

training,

6.17, p= 0.035);

Computerized

four sweeteners

ysis software Guelph, zontal

( CSAr,,TM ) (CSA V4.3;

Canada)

tion of either line

horizontal

Analysis Temporal

was modified

a vertical scale

Profile

Anal-

Compusense

Inc,

to make

the presenta-

line scale or the existing

possible.

Both

the

vertical

horiand

lines were 60 pixels long and were labelled

with the anchors ‘not sweet’ and ‘very sweet’. The time-intensity software (CSATpATM)was programmed to collect responses every 0.5 s for a total of 60 s (120 records). The panelists were instructed to drink the

tation,

there

ratings

horizontal

for four

of the

parameters: ZMAX (F( 1,9) = 16,57, (F(1,9) = 12.24, p = O-007); DEC

= 5.59, p = 0.042);

samples of sweetener similar to those used in testing were presented to the panelists for evaluation. The Sensory

DISCUSSION

DECAREA

see Fig. 2. When averaged

were significant

differences

(F(1,9)

=

across orienamongst

for three of the eight time-intensity

the pa-

rameters:

AUC(F(3,27) = 3.88, p = 0.020); INCANGLE (3,27) = 4.45, p = 0.011); DECAREA (F(3,27) = 4.09, p = O-016); see Fig. 3. Representative

time-intensity

curves

for horizontal

and vertical orientation for each sweetener are shown in Fig. 4. Average curves produced by simple arithmetic averaging are skewed toward the panelist with the longest duration and do not provide an accurate illustration of the individual parameters. For this reason,

sample through the straw, hold it in their mouths for 3 s and then swallow. Time-intensity evaluations were started immediately upon ingestion to capture the time

the curves in Fig. 4 are the time-intensity curves from one panelist who is in the middle of the range. These curves graphically illustrate the individual time-inten-

course

sity parameters.

of maximum

intensity.

The

panelists

used

a

Time-Intensity Perqbtion of Sweetener Solutions

INC ANGLE

+

123

DECANGLE

DEC AREA

FIG.

1. Time-intensity

TABLE

curve and parameters.

1. Time-intensity

Parameters

and their Definitions

Parameter Maximum

intensity

Time to maximum Duration

Abbreviation

Definition

ZMAX

The maximum sweetness intensity (up to 60 pixels) of each sample. The time (in seconds) at maximum intensity. The time (in seconds) for sweetness perception (from first perception to the end of the perception). The angle of increase to maximum intensity. This can be interpreted to be the rate of onset of sweetness of the sample. The area under the increasing portion of the curve. The angle of decrease from maximum intensity. This can be interpreted to be the rate of decrease of sweetness perception. The area under the decreasing portion of the curve. The total area under the time-intensity curve.

TMAX DUR

intensity

Increase

angle

ZNC ANGIE

Increase Decrease

area angle

ZNCAREA DEC ANGLE DEC AREA AUC

Decrease area Area Under the Curve

450 T

30

65

I

300 -

T

400 c

29

250 350 _c

28 300 4 27

250 :

: s

200

26

$

3 G

150

200 I 150

25 100 24

61

I

60 I

23 DANGLE

AUC

FIG. 2. Mean values for significant

time-intensity

parameters

affected

by orientation.

DAREA

i

OHorizontd

8Vertical

1

124

L. M. Duizeq K. Bloom, C. J. Findlay 82 _

500 450

81

400 350

80

300 3

200 =

250

i

;

9 19

2

I50

200 78

150 100 50 0

76 INC

AUC

DEC

ANGLE

AREA

FIG. 3. Mean values for significant time-intensity parameters affected by sweetener.

50 -

W

(a)

40 ..

.: 0

5

10

:

I5

!

:

:

20

:

:

25

Time Iatcrvd

:

:

30

:

:

35

:

40

.,,,

I 45

0

5

10

,,,,,,,,,( 15

20

25

30

35

40

45

Time Intern1 (I)

(8)

50 -

50 @I

@I 40 -.

40 -P

: : : 0

5

10

I5

:

20

:

: 25

Time lntervd

FIG. 4. Time-intensity

:

: 30

:

: 35

: 40

:

I 45

(s)

:::::: 0

5

10

IS

:::;::* 20

25

30

35

40

45

Time Intcrvd (I)

curves from one individual for each sweetener: (a) sucralose; (b) aspartame; (c) acesulfame k; (d) sucrose.

Multiple comparisons with repeated-measures t-tests were conducted to isolate the differences between sweeteners for the three significant parameters. The results indicated that the area under the curve was significantly greater for sucralose as compared with aspartame (t(9) = 2.61, p = O-028)) and with sucrose (t(9) = 2.32, p= 0.046). Lik ewise, the area under the curve was significantly greater for acesulfame k compared with aspartame (t(9) = 2.80, p = 0.021)) and with sucrose (t(9) = 2.31, p = 0.046). The angle of increasing intensity was significantly greater for acesulfame k as compared with sucralose (t(9) = 3.31, p = 0.009), and with aspartame (t(9) = 4.01, p = 0.003). Finally, the area under the

decreasing portion of the curve was significantly greater for sucralose as compared with aspartame (t(9) = 3.04, p = O-014), and with sucrose (t(9) = 2.51, p = 0.033), and greater for acesulfame k as compared with aspartame (t(9) = 2.64, p = 0.027). There were no significant Orientation X Sweetener interaction effects for any of the eight parameters (Table 2). This means that differences in perceptions of the four sweeteners (sucralose, acesulfame k, aspartame, and sucrose) were not affected by the orientation of the time-intensity scale. These results provide evidence that both horizontal and vertical scales can be used to assess time-intensity sweetness responses. In addition, the lack of significant

ofSweetener Solutions

Time-Intensity Per@&m differences

between

horizontal

and vertical

scales for

among panelists to homogenize

orientation

the parameters of TMAX, INC ANGLE, INC AREA and DUR provide support for the use of either horizontal or

Results from this study, regardless were similar to those obtained by other

vertical

dependent

time-intensity

It has been affects

the

area

significantly

scales.

demonstrated under

that maximum

the

curve,

high correlations

the significant observed

by

the IMAX and

1992).

In this study

AUC, DEC A??EA and DEC ANGLE results

between

the two orientations

with the significant

1A4AXvalues

Lawless & Clark (1992) evaluation

as indicated

between

the AUC and DEC AREA (Duizer,

intensity

have defined

as something

because Noble

temporal be

to be

greater

line.

Maximum

the vertical that rated

bias. Panelists’

sweetness

intensity

(IMAX)

2). Perhaps

tally. A vertical

on than

These

scale

(Fig.

(1991)

groups

for

movement

strength

of the arm requires

and

hence,

reducing

mouse

along the vertical

not

for

the precision

the use

line. In contrast,

less strength The shoulder

1991).

left

This movement

trolled excursion The

motor

to

is required

of the re-

to move the

and wrist are rotated

right

movement

produces

a smaller,

observed

tions may be minimized

between

to reduce

the

to

(Chaffin, better

of the mouse on the horizontal

differences

control,

of the mouse along the vertical

arm horizontally.

the

fine

in the movement

line. This lack of precision

sults in a larger excursion

accommodate

con-

line.

two orienta-

possible

bias due to

tion on the vertical and horizontal

lines may increase

the

precision of the panel (O’Mahony & Wong, 1989). Further training with reference samples might allow the panelists to become

more familiar with both orientations.

orientation

can

be counterbalanced

Finally,

within

and

Orientation

IMAX

TMAX DUR AUC INCANGLE INCAREA DECANGLE DECAREA (N.S.=p>0.05)

for acesulfame

k indicates

perception

that acesulfame aspartame

by

significantly of acesul-

and sucralose. by Ott et d.,

k had a faster

and alitame.

Time-intensity evaluations may be performed horizontal and vertical orientations. Although the time-intensity ferent,

0.003 N.S. N.S. 0.007 N.S. N.S. 0.042 0.035

Sweetener

Orientation X Sweetener

N.S. N.S.

N.S. N.S. N.S.

N.S.

0.016 0.011 N.S. N.S. 0.018

N.S. N.S.

N.S. N.S. N.S.

parameters

the muscle

groups

zontally compared These

findings

support

scales

difverti-

may be due to the mouse

the use of both

to collect

attributes

will provide

dependent

to move

is greater

hori-

to vertically.

for two different opment

intensity

This difference

required

in both most of

are not significantly

the value for maximum

cally than horizontally.

changes

horizontal

time-intensity simultaneously.

a new

tool

responses This devel-

to investigate

in foods and consumer

time

products.

ACKNOWLEDGEMENTS Financial

support

the Industrial tional

TABLE 2. Summary of pvalues for Sweetener and Orientation Parameter

intensity

The

CONCLUSIONS

and vertical

orientation of the line. For example, the presentation of reference samples to anchor responses in the same posi-

scale

maximum

results were similar to those observed

rate of onset than sucrose,

max’

which was obtained

than that of aspartame

who concluded

of the

the mouse vertically versus horizon-

of the biceps, brachialis and triceps muscles which are the larger muscles in the arm. These larger muscles are used

of onset to the ‘rate

intensity.

INC ANGLE observed

fame k is quicker

higher

muscle

the rate

to reach

maximum

affects

in fNC ANGLE

is similar (1990)

of

must

concentration

reflects

measure

the height to

sucrose.

sweeteners

Differences

that the rate of onset of sweetness

13%

this is due to the different

involved in moving

rated

time

time-intensity

scale was approximately on the horizontal

responses

of the time-intensity

sweetener

by Ott & Palmer

by dividing

entation

as

This

to 9%

measurement,

sweeteners

reported

equisweet

parameters.

sweeteners.

dif-

for the four sweeteners,

have stated that for the purpose

perception

equisweet,

the

by the orientation

were

amongst

2, Fig. 2). a bias in sensory

which causes a response

could be a potential

the

samples

effects.

of orientation, researchers. In-

there were no significant

intensity

et al. (1991)

an inaccurate reflection of intensities. The results from this study suggests that for certain parameters line oriwere affected

of orientation, in maximum

time-intensity

are concomitant

(Table

ferences

125

Research

for this study was provided

Research

Assistance

Council

Canada.

not have been completed

without

Program

in part by of the Na-

This research the generous

could cooper-

ation of our panelists.

REFERENCES Birch, G. G. & Munton, S. L. (1981). Use of the ‘Smurf’ in taste analysis. Chem. Senses, 6,45-52. ChafYin, D. B. (1991). &~~puti~na~Biomchanics.John Wiley & Sons, Inc., Toronto, p. 397. Cliff, M. & Heymann, H. (1993). Development and use of time-intensity methodology for sensory evaluation: A review. FoodResearchInternational, 26, 375-385.

126

L. M. Duizer, K Bloom, C.J. Findlay

Dubois, G. E. & Lee, J. F. (1983). A simple technique for the evaluation of temporal taste properties. Chem.Senses,7,237-247. Duizer, L. M. (1992). An assessment of sensory time-intensity and instrumental methods for bovine muscle tenderness measurements. M.Sc. Thesis. University of Guelph, Guelph, ON. Guinard, J-X., Pangborn, R. M. and Shoemaker, C. F. (1985). Computerized procedure for time-intensity sensory measurements. Journal ofFood Science, 50,543-546. Larson-Powers, N. & Pangborn, R. M. (1978). Paired comparison and time-intensity measurements of the sensory prop erties of beverages and gelatins containing sucrose or synthetic sweeteners. Journal ofFood Science, 43,41-46. Lawless, H. T. & Clark, C. C. (1991). Psychological Biases in time-intensity scaling. Food Technology, 46, 81,84-86,90. Lee, W. E. (1985). Evaluation of time-intensity sensory responses using a personal computer. Journal of Food Science, 50,1750-1753. Neilson, A. J. (1957). Time-intensity studies. Drug and Cosm. Indusby, 80,88-93.

Noble, A. C., Matysiak, N. L. & Bonnans, S. (1991). Factors affecting the time-intensity parameters of sweetness. Food Technology, 45, 121-126. O’Mahony, M. & Wong, S-Y. (1989). Time-intensity scaling with judges trained to use a calibrated scale: Adaptation, salty and umami tastes. Journal of Sensq Studies, 3, 217-236. Ott, D. B., Edwards, C. L. & Palmer, S. J. (1991). Perceived taste intensity and duration of nutritive and non-nutritive sweeteners in water using time-intensity evaluations. Journal ofFood Science, 56,535-542. Ott, D. B. & Palmer, S. J. (1990). Ingestion and expectoration sampling methods of four tastes in a model system using time-intensity evaluations. Journal of Sensmy Studies, 13, 53-70. Schmitt, D. J., Thompson, L. J., Malek, D. M. & Munroe, J. H. (1984). An improved method for evaluating time-intensity data. Journal of Sensory Studies, 49, 539-542,580. Schwartz, M. (1980). Sensory screening of synthetic sweeteners using time-intensity evaluation. Journal of Food Science, 45,577-581.