Cover Image
close this bookResearch Methods in Nutritional Anthropology (UNU, 1989, 201 pages)
close this folder3. Methodological procedures for analysing energy expenditure
View the document(introductory text...)
View the documentIntroduction
View the documentSurvey of habitual activities
View the documentDetermination of critical activities
View the documentDetermination of key participants
View the documentMeasuring energy expenditure rates
View the documentTime-motion analysis
View the documentEstimation of energy expenditure rates from time-allocation data
View the documentAssessment of endurance capacity
View the documentSummary
View the documentReferences

Time-motion analysis

Time-motion analysis is a standard analytic procedure for determining the time and energy spent in activities over a period of time (see Gross, 1984). A data-collection period of one to two weeks is considered ideal, although three days, including one weekend day (in populations which have a defined work week), is usually acceptable. In addition, it is necessary to repeat the data collection process at appropriate times in a season or annual cycle where significant shifts in energy expenditure patterns are to be expected.

A number of methods exist for recording the amount of time spent by an individual in various activities. All of these require that complex activities be broken down into work elements for which some means of assessing level of exertion can be established. A work element is defined as "an activity of relatively constant energy cost and of characteristic motion and composition for an individual under specific conditions" (Consolazio et al., 1963, p. 327).

Such elements are easily defined verbally and by observation; for example, washing dishes, digging a ditch, typing, or sweeping a floor would all be considered "work elements." Table 6 describes the individual work elements involved in different stages of potato production by Nuñoa men (Thomas, 1973a). Time expenditure is calculated in terms of inputs into a 500 m² potato plot, which is a common unit of production. Although the number of separate classifications used in recording and analysis will depend upon the precision desired, Weiner and Lourie (1969) note that 15 activity categories are usually sufficient for most purposes. It should be kept in mind that while more specific categories can later be grouped into more general ones for analytic purposes, the reverse is not true.

The traditional method for obtaining an activity record for time-motion analysis is through direct observation of individuals. Observers use a recording sheet and stopwatch, although any watch with a second hand is usually adequate. Tape-recorders or communication through use of a walkie-talkie radio set with a central recording station where observations are coded has also been utilized. Although this technique is highly accurate, its practicability is questionable under most conditions. Except in situations where many subjects are engaged in similar activities in one location and can, therefore, be monitored as a group, the number of personnel required for direct simultaneous observation is exceptionally large, and the number of subjects who can be monitored is quite small.

A more practical method is the diary technique, which requires the subject to keep a record of all activities. The diary technique can be used alone or in conjunction with the continuous observation method, especially in instances where people spend part of the day together (as at work) and another part separately. Simple code letters are devised for each type of activity and the subject records his/her activities for each minute or at five-minute intervals, depending on the precision required. Figure 3 presents an example of a diary coding sheet for which a minute-by-minute record is to be kept.

Table 6. Estimated energy-expenditure rates for activities in different stages of potato production in Nuñoaa

Activity complex Energy expenditure
rate (kcal/min)
Time expenditure (min/500 m²) Energy cost (kcal/500 m²)
I. Field preparation and planting  
Stone removal 3.5 95 332.5
Irrigating 6.5 215 1,397.5
Foot ploughing (2 men) 6.3 476 2,998.8
Foot ploughing (wife) 3.1 236 731.1
Breaking up clods 4.5 167 751.5
Spreading dung 4.5 357 1,606.5
Planting 6.0 217 1,302.0
Planting (wife) 3.3 217 716.0
Walking to and from fields 5.0 480 2,400.0
Transporting dung 6.0 300 1,800.0
Transporting dung (wife) 4.5 300 1,350.0
Transporting seed 5.5 80 440.0
Transporting seed (wife) 4.5 80 360.0
Subtotal   3,200 16,185.9
Percentage of total   41.3 47.0
2. Weeding, ridging  
Walking 3.5 200 700.0
  4.5 80 360.0
Subtotal   280 1,060.0
Percentage of total   3.6 3.1
3. Harvest  
Picking 4.2 1,667 7,043.4
Walking to and from fields 5.0 400 2,000.0
Transporting harvest 5.5 120 660.0
Transporting harvest (wife) 4.5 120 540.0
Subtotal   2,307 10,243.4
Percentage of total   29.6 29.8
4. Food preparation  
Sorting potatoes 3.5 240 840.0
Making chuño 3.5 1,500 5,250.0
Seed storage 3.5 240 840.0
Subtotal   1,980 6,930.0
Percentage of total   25.5 20.1
Total for potatoes   7,787 34,419.3

a. Energy expenditure rates are calculated for an individual man unless otherwise indicated Source: Thomas, 1973a, p. 77.

Despite its ease in administering, the diary technique requires that subjects be literate and sufficiently motivated. It is necessary to check their recordings at least once a day to make sure that all time is accounted for and to advise them concerning any difficulties that may be encountered in recording. In all, if carefully monitored, this method has been found to offer adequate precision for most purposes, while allowing more subjects to be investigated than does the direct observation method.

Blair (1980) has overcome the problem of having to write down activities, and hence interfering with habitual activity. Subjects are given a pocket calculator-sized (battery-powered) device with 10 numbered keys to correspond with 10 groups of normally performed activities. As a new activity begins, the subject presses the appropriate key which issues a tone sequence. This sequence is fed into an Avionics two-channel electrocardio - recorder and can be recorded simultaneously with heart rate.

For non-literate groups, where the diary technique cannot be employed, it remains the task of the investigator to observe and record activities. To overcome the sampling limitations of direct observation, Johnson (1975, 1978) has devised the "randomized household visit method' of determining time allocation for populations and population segments. In his study on the Machiguenga Indians of lowland Peru, he made 2,500 behavioural observations during one year on a large number of individuals.



Fig. 3. Diary coding sheet (after Weiner and Lourie, 1969, p. 279)

Table 7. Mean daily time allocations of married adult a

 

Women (N = 20)

Men (N = 15)

Major activity category,and subcategory

Min/day

%

Min/day

%

Eating

71

4.9

55

3.8

Food preparationb

12

0.8

141

9.8

Child-rearingb

-

0.1

69

4.8

Manufacture

81

5.6

124

8.6

Woodworkb

52

3.6

5

0.3

Cotton clothb

1

0.1

105

7.3

Other

28

1.9

14

1.0

Wild foodsb

122

8.5

51

3.5

Collecting

22

1.5

19

1.3

Fishing

45

3.1

18

1.3

Huntingb

45

3.1

0

0

Other

10

0.7

14

1.0

Garden labourb

144

10.0

51

3.5

Clear, burn, plantb

29

2.0

0

0

Weedingb

45

3.1

2

0.1

Harvesting

47

3.3

39

2.7

Other

23

1.6

9

0.6

Idle

141

9.8

149

10.3

Hygiene

20

1.4

35

2.4

Visiting

62

4.3

45

3.1

Other

126

8.8

59

4.1

Daylight hoursc

779

54.2

779

53.9

Night hours

660

45.8

660

45.8

Totalc

1,439

100.0

1,439

99.7

a. Adapted from Johnson, 1975, p. 308.
b. Differences between sexes significant at p <: .01 (t-test).
c. Totals vary slightly from expected due to rounding.

Observation times were statistically randomized according to season and time of day throughout an annual cycle, with written descriptions of each person's activities being recorded at the moment of observation. These descriptions were then transformed into activity codes for computer processing. From the total combined observations, estimates were computed of average time-expenditure patterns for men and women in the population. Some of the results of this analysis are shown in table 7 where mean daily time allocations for married adult males and females are presented. With this approach, only minimum participant co-operation is necessary, and sampling of large numbers of individuals over long periods of time is possible. Whereas random activity sampling is relatively easy to conduct, takes comparatively little time, and covers a representative range of habitual activites, it is not a substitute for intensive studies of productive activities, where reconstructing the productive process is of importance.

A final method of determining time allocation is through the use of questionnaires that ask subjects to recall their activities for previous days, weeks or even longer periods of time. "Selfassessment" questionnaires may also ask the subject to estimate his/her level of activity in various activities. Levels of activity are usually given classifications such as very active, fairly active, average, fairly inactive, and very inactive. The meaning of these categories is not explained to the subjects: they are simply told to use their own judgment. The reliability of questionnaire methods has been found lacking, especially in that subjects tend to exaggerate the duration and/or intensity of their physical activities. Nevertheless, questionnaires are often used and, if employed with care and a certain of degree of scepticism, can yield relative impressions that are adequate for many research purposes.

Name: Occupation:
Serial number: Ethnic group of nationality:
Age: Place of examination:
Sex: M/F Date of this record:
Observer's name: Day of week:

Time-budget of daily activities      
1. Night's rest: (mins.) 5. Physical effort during work:  
    light (2-3 kcal/min)- (mins.)
2. Bathing, dressing, eating: (mins.) moderate (3-4 kcal/min)- (mins.)
    heavy (4-5 kcal/min)- (mins.)
3. Transportation to and from work:      
walking- (mins.) 6. Leisure time posture:  
driving private car- (mins.) sitting- (mins.)
travelling by bus or rail- (mins.) standing- (mins.)
cycling- (mins.) walking- (mins.)
4. Work posture:   7. Physical effort during leisure time:  
sitting- (mins.) light (2-3 kcal/min)- (mins.)
standing- (mins.) moderate (3-4 kcal/min)- (mins.)
walking- (mins.) heavy (4-5 kcal/min)- (mins.)

Note should be taken of time, number and total duration of all peak loads in day.

Fig. 4. 24-hour self-assessment recall questionnaire (after Weiner and Lourie, 1969, p. 280).

Figure 4 provides a sample 24-hour self-assessment recall questionnaire. In it, the subject is asked to recall the time spent sleeping, bathing, dressing and eating, travelling to and from work, working, and engaging in leisure activities. For work and leisure activities, further note is made of the amount of time spent in various "postures" (i.e. sitting, standing, and walking) and in various degrees of physical exertion (i.e. light, moderate, and heavy). In this manner very general information on activity patterns and levels of activity may be derived.

Nag et al. (1978) used a similar method in determining time inputs for a variety of productive activities for different age and sex groups in a Nepalese and a Javanese village. A number of households in each village were visited every day at regular intervals during time-spans ranging from six months to over a year. Upon each visit, each household member was asked to recall his/her activities during the previous 24-hour period. Fourteen activity categories were used including child care, agriculture, wage labour, and food preparation. Average-time inputs into various types of activities were then calculated and compared between villages and by age/sex categories. Some of the results of this study are shown in table 8.

Table 8. Average time input (in hours) per person per day in different work activities among the males of various age groups in a Javanese village

Age group and sample size

 

6-8

9-11

12-14

15-19

20-29

30-39

40-49

50+

Activity

(6)

(7)

( 10)

(6)

(5)

(9)

(8)

(3)

Child care

1.2

0.5

0.3

0.0

0.0

1.0

0.2

0.0

Household food preparation

0.0

0.0

0.1

0.0

0.1

0.1

0.1

0.1

Firewood collection

0.6

0.8

0.9

0.2

0.1

0.3

0.1

0.2

Other household maintenance work

0.1

0.1

0.2

0.1

0.2

0.1

0.1

0.0

Animal care

1.7

1.5

2.5

1.6

2.4

1.3

0.9

0.8

Wage labour (agricultural)

0.0

0.0

0.1

0.3

0.7

0.4

0.1

0.0

Wage labour (non agricultural)

0.0

0.0

0.0

1.4

1.0

1.7

0.7

0.0

Handicrafts

0.0

0.0

0.1

0.9

0.1

0.1

0.6

0.5

Reciprocal labour exchange

0.0

0.0

0.1

0.7

1.0

0.8

0.6

0.6

Irrigated rice cultivation (own land)

0.0

0.0

0.2

1.1

2.2

1.4

2.4

3.0

Garden cultivation (own land)

0.0

0.1

0.1

0.2

0.1

1.1

0.9

1.0

Trading

0.0

0.0

0.0

1.1

0.3

0.6

1.1

0.0

Preparation of food for sale 0.0

0.0

0.0

0.0

0.0

0.3

0.6

0.7

 
Other

0.0

0.1

0.1

0.3

0.5

0.2

0.3

0.4

Total for household maintenance (1-4)

1.9

1.4

1.5

0.3

0.4

1.5

0.5

0.3

Total for directly productive work (5-14)

1.7

1.7

3.2

7.6

8.3

7.9

8.2

7.0

Total for all work (l-14)

3.6

3.1

4.7

7.9

8.7

9.4

8.7

7.3

Source: Nag et al., 1978, p. 294.

For many purposes, consideration of time expenditure alone is sufficient, and no new information is gained by estimating energy expenditure. This is particularly the case for a sedentary life-style where activities show little variation in energy expenditure rate. Conversely, where rates show considerable variation between tasks and this is not reflected in time inputs, energy estimates provide valuable supplementary data. For instance, table 9 compares time and energy inputs of a 12-year-old Nuñoa boy and his father performing a typical day's herding (Thomas, 1973a). Whereas time spent and work performed are identical, the lighter boy spends over 30 per cent less energy in completing the activity. When these data are interpreted in terms of food energy saved in the course of a year as a consequence of the child's herding, this energy saving comes to about 107,000 kcals or the food-calorie equivalent of almost six sheep.