Levine (2001) & Moray (1959)

Helping behaviour refers to voluntary actions intended to help others and is a form of prosocial behaviour. Theories about helping behaviour include: 

Kin selection theory: this refers to the tendency to perform behaviours that may favour chance of survival of people with a similar genetic base (Hoffman 1981)

Reciprocal altruism: this holds that the incentive for an individual to help in the present is based on the expectation of the potential receipt in the future (Trivers 1971) 

Responsibility-prosocial value orientation: holds that a strong influence of helping behaviours is a feeling and belief in one's responsibility to help, especially when combined with the belief that one is able to help the other person (Staub, 2003) 

Social exchange theory: people help because they want to gain goods from the one being helped. They calculate rewards and costs of helping others, aiming to maximise the rewards and  minimise the costs (Foa 1975)

Milgram (1970) proposed that people in urban areas are less helpful than those in rural areas because they experience greater sensory overload: urban dwellers restrict their attention mainly to personally relevant events. Strangers and their situations of need may, therefore, go unnoticed. 

Studies conducted in several different countries (including USA, Saudi Arabia and Sudan) have found that people living in urban areas tend to be less helpful than those in rural settings (Hedge and Yousif). Virtually all of the studies of community differences in helping have focused on the single variable of population size, most often testing the hypothesis that the tendency to help strangers declines as the size of the city population increases. Steblay (1987) found general support for this hypothesis with the decline in helping rate beginning at populations of 300,000. She also found that urban environments of 300,000 people or more and rural environments of 5,000 people or less were the worst places if one was looking for help.

A major cultural difference in helping behaviour is the difference between collectivism and individualism. Collectivists attend more to the needs and goals of the group they belong to, and individualists focus on their own selves. Therefore, collectivists would be more likely to help ingroup members, but less frequent than individuals to help strangers (Triandis, 1995). 

• Although many studies have demonstrated that helping rates differ between communities in a single country, almost no systematic cross- cultural research of helping behaviour had been conducted prior to this study.

• The aim of this study was therefore to look at helping behaviour, in a wide range of cultures, in large cities around the world in relation to four specific community variables:

  (i) Population size
  (ii) Economic well-being
  (iii) Cultural values (individualism-collectivism, simpatia) (iv) Walking speed (pace of life). 

• This study had three main goals:

• (i) To determine if a city’s tendency to offer non-emergency help to strangers is stable across situations over a wide range of cultures i.e. is helping strangers a cross-culturally meaningful characteristic of a place?

• (ii) To obtain a descriptive body of data on helping behaviour across cultures using identical procedures i.e. does helping strangers vary cross- culturally?

  (iii) To identify country-level variables that might relate to differences in helping i.e. what are some community characteristics that are related to helping of strangers across cultures?

• Three overlapping theoretical explanations for community-level differences in helping behaviour, none of which had been previously considered in cross-cultural research, were tested:  (i) Economic explanations  (ii) Cultural values (iii) Cognitive explanations: pace of life. 

• This was a field experiment that used an independent measures design.

• The eld situation was 23 large cities around the world including Rio de Janeiro (Brazil), Calcutta (India), Madrid (Spain), Shanghai (China), Budapest (Hungary), Rome (Italy), New York (USA) and Kuala Lampur (Malaysia).

• The experiment had three independent variables (IVs):
  (i) whether the victim dropped a pen
  (ii) whether the victim had a hurt/injured leg
  (iii) whether the victim was blind and trying to cross the street.

• The dependent variable (DV) was the helping rate of the 23 individual cities (calculated to give each city an Overall Helping Index).

• The three measures of helping were correlated with statistics reflecting population size, economic well-being, cultural values (individualism- collectivism, simpatia) and the pace of life for each of the 23 locations. 

• Participants in this study were large cities in each of 23 countries – in most cases the largest in each country i.e. individuals in each of these cities at the time of the experiment.

• Each of the three helping measures and the walking speed measure were administered in two or more locations, in main downtown areas, during main business hours, on clear days, during the summer months of one or more years between 1992 and 1997.

• For the dropped pen and hurt leg situations, only individuals walking alone were selected. Children (younger than 17 years old), and people who were physically disabled, very old, carrying packages etc (i.e. those who might not be fully capable or expected to help) were excluded.

• Participants were selected by approaching the second potential person who crossed a predetermined line 

• Data was collected by either interested, responsible students who were either travelling to foreign countries or returning to their home countries for the summer, or by cross-cultural psychologists and their students in other countries who volunteered to assist the authors.

• All experimenters were college age and dressed neatly and casually. To control for experimenter gender e ects and to avoid potential problems in some cities, all experimenters were men.

• To ensure standardisation in scoring and to minimise experimenter effects:

  (i) all experimenters received both a detailed instruction sheet and on-site field training for acting their roles, learning the procedure for participant selection and scoring of participants

  (ii) the experimenters practised together

  (ii) no verbal communication was required of the experimenter. 

1. (i)  Dropped pen. Walking at a carefully practiced, moderate pace (15 paces/10 seconds), experimenters walked toward a solitary pedestrian passing in the opposite direction. When 10 to 15 feet from the participant, the experimenter reached into his pocket and accidentally, without appearing to notice, dropped his pen behind him, in full view of the participant, and continued walking past the participant. A total of 214 men and 210 women were approached. Participants were scored as having helped if they called back to the experimenter that he had dropped the pen and/or picked up the pen and brought it to the experimenter. 

1. (ii)  Hurt leg. Walking with a heavy limp and wearing a large and clearly visible leg brace, experimenters accidentally dropped and unsuccessfully struggled to reach down for a pile of magazines as they came within 20 feet of a passing pedestrian. A total of 253 men and 240 women were approached. Helping was de ned as o ering to help and/or beginning to help without o ering. 

1. (iii)  Helping a blind person across the street. Experimenters, dressed in dark glasses and carrying white canes, acted the role of a blind person needing help getting across the street. (The canes and training for the role were provided by the Fresno Friendship Centre for the Blind.) Experimenters attempted to locate downtown corners with crosswalks, tra c signals, and moderate, steady pedestrian ow. They stepped up to the corner just before the light turned green, held out their cane, and waited until someone o ered help. A trial was terminated after 60 seconds or when the light turned red, whichever occurred rst, after which the experimenter walked away from the corner. A total of 281 trials were conducted. Helping was scored if participants, at a minimum, informed the experimenter that the light was green. 

• Two of the three correlations were significant. All three intercorrelations were in the positive direction.

• No significant gender differences in helping behaviour were found in the two conditions in which relatively equal numbers of male and female participants were targeted by the experimenter (hurt leg, dropped pen): dropped pen, M (men) = .67, M (women) = .69, t(22) = .39, ns; hurt leg, M (men) = .63, M (women) = .65, t(22) = .75, ns.

•  An Overall Helping Index was calculated, combining results for the three helping measures. Results showed that the most helpful cities/ countries were (1) Rio de Janeiro (Brazil), 93%, (2) San Jose (Costa Rica), 91% (3) Lilongwe (Malawi), 86%. The least helpful cities/countries were (23) Kuala Lampur (Malaysia), 40% (22) New York (USA), 45%, (21) Singapore (Singapore), 48%. 

• On average, there were low correlations between the community variables and helping measures. The only statistically reliable relationship was between the economic productivity measure and overall helping: cities that were more helpful tended to have lower Purchasing Power Parity (PPP). Although statistically insignificant, there was a small relationship between walking speed and overall helping, with participants in faster cities somewhat less likely to help.More individualistic countries showed somewhat less overall helping and less helping in the hurt leg situation than collectivist countries, but none of the correlations reached a significant level. There was no relationship between population size and helping behaviour. The two community variables of economic productivity and individualism-collectivism and walking speed were highly intercorrelated: economic productivity was positively correlated with individualism and negatively correlated with walking speed i.e. faster paces had stronger economic productivity. Individualism was also negatively correlated with walking speed i.e. faster places were more individualistic. Simpatia countries (Brazil, Costa Rica, El Salvador, Mexico and Spain) were, on average, more helpful than non simpatia countries. Overall, a city’s helping rate was relatively stable across all three measures. 

• The helping of strangers is a cross-culturally meaningful characteristic of a place.

• There are large cross-cultural variations in helping rates.

• Helping across cultures is inversely related to a country’s economic productivity.

• Countries with the cultural tradition of simpatia are, on average, more helpful than countries with no such tradition.

• Although faster cities tend to be less helpful than slower cities, the link between economic health and helping is not a by-product of a fast pace of life in affluent societies.

• The value of collectivism-individualism is unrelated to helping behaviours. 

• Attention has a selective nature.

• Broadbent (1958) argued that the world is composed of many more sensations than can be handled by the perceptual and cognitive capabilities of the human observer. To cope with the flood of available information, humans must selectively attend to only some information and somehow ‘tune out’ the rest. Attention, therefore, is the result of a limited-capacity information-processing system.

• Almost all early models of attention assumed serial processing, a step-by-step process in which each operation is carried out in turn. The first of these was Broadbent’s (1958) filter model, followed by Treisman’s (1964) attention model and the pertinence model (Deutsch & Deutsch, 1963; Norman, 1969).

• Early attempts to explain divided attention, such as Kahneman’s (1973) central capacity theory, also assumed serial processing.

• However, Allport’s (1980b) multi-channel theory of divided attention holds that two or more operations can be carried out at the same time.

• There are two main methods of studying attention:

  1. (i)  Selective attention – here people are presented with two or more simultaneous ‘messages’, and are instructed to process and respond to only one of them. The most popular way of doing this is to use shadowing in which one message is fed into the left ear and a different message into the right ear (through headphones). Participants have to repeat one of these messages aloud as they hear it. The shadowing technique is a form of dichotic listening which was first used by Cherry (1953) when he studied the cocktail party phenomenon.

  2. (ii)  Divided attention – this is a dual-task technique in which people are asked to attend and respond to both (or all) the messages. Whereas shadowing focuses attention on a particular message, the dual-task method deliberately divides people’s attention. 

• Cherry’s (1953) method of ‘shadowing’ one of two dichotic messages for his study of attention in listening found participants who shadowed a message presented to one ear were ignorant of the content of a message simultaneously presented to the other ear.

• Other researchers then moved on from Cherry’s work on how people can attend to one message by investigating why so little seemed to be remembered about the other conversations (Hampton & Morris, 1996).

• The first experiment in this study aimed to test Cherry’s findings more rigorously whilst the second and third experiments aimed to investigate other factors that can affect attention in dichotic listening. 

• All tasks were laboratory-based, had high levels of control and had an IV and DV. Therefore they were all laboratory experiments.

• In all tasks, the apparatus used was a Brenell Mark IV stereophonic tape recorder modified with two amplifiers to give two independent outputs through attenuators, one output going to each of the earpieces of a pair of headphones. Matching for loudness was approximate, by asking participants to say when two messages that seemed equally loud to the experimenter were subjectively equal to them. 

1. Experiment 1

   This used a repeated measures design. This independent variables (IVs) were:

   (i) the dichotic listening test

   (ii) the recognition test
   The dependent variable (DV) was: the number of words recognised correctly in the rejected message. 

1. Experiment 2
   This used an independent measures design.
   The independent variable (IV) was: whether or not instructions were prefixed by the participant’s own name. The dependent variable (DV) was: the number of effective instructions
   

1. Experiment 3
   This also used an independent measures design.
   The independent variables (IVs) were:

   (i) whether digits were inserted into both messages or only one

   (ii) whether participants had to answer questions about the shadowed message at the end of each passage or whether participants had to merely remember all the numbers s/he could.

   The dependent variable (DV) was: the number of digits correctly reported. 

• Participants were undergraduates and research workers of both sexes.

• Participant numbers are not given for Experiment 1 but 12 participants took part in the experimental conditions in Experiment 2 and two groups of 14 participants were used in Experiment 3. 

 Outline of the procedure/study

• Before each experiment the participants were given four passages of prose to shadow for practice. All passages throughout the study were recorded by one male speaker.

Experiment 1

• A short list of simple words was repeatedly presented to one of the participant’s ears whilst they shadowed a prose message presented to the other ear. (The word list was faded in after shadowing had begun, and was equal in intensity to the shadowed message. At the end of the prose passage it was faded out so as to become inaudible as the prose finished.)

• The word list was repeated 35 times.

• The participant was then asked to report all he could of the content of the rejected message.

• S/he was then given a recognition test using similar material, present in neither the list nor the passage, as a control.

• The gap between the end of shadowing and the beginning of the recognition test was about 30 seconds. 

1. Experiment 2

   • This experiment was conducted to nd out the limits of the efficiency of the attentional block.

   • Participants shadowed ten short passages of light fiction.

   • They were told that their responses would be recorded and that the object of the experiment was for them to try to score as few mistakes as possible.

   • In some of the passages instructions were interpolated, but in two instances the participants were not warned of these.

   • In half of the cases with instructions these were prefixed by the participant’s own name.

   • The ‘no instructions’ passages were interpolated in the table at random.

   • The passages were read in a steady monotone voice at about 130 words per minute.

   • Participants’ responses were tape-recorded and later analysed. 

1. Experiment 3

   • Experiment 2 indicated that instructions might alter the set of instructions a participant in such a way as to alter the chances of material in the rejected message being perceived. Experiment 3 tested this point further.

   • Two groups of 14 participants shadowed one of two simultaneous dichotic messages.

   • In some of the messages digits were interpolated towards the end of the message. These were sometimes present in both messages, sometimes only in one. The position of the numbers in the message and relative to each other in the two messages were varied, and controls with no numbers were also used, randomly inserted.

   • One group of participants was told that it would be asked questions about the content of the shadowed message at the end of each message, the other group was speci cally instructed to remember all the numbers that it could. 

• Words presented in shadowed message: 4.9 out of 7
• Words presented in rejected message: 1.9 out of 7
• Words presented for the first time in recognition test: 2.6 out of 7
  
• (mean number of words recognised) 

• -  There was no trace of material from the rejected message being recognised.

• -  The difference between the new material and that from the shadowed message was signi cant at the 1 per cent. level.

• - The 30-second delay was unlikely to have caused the rejected material to be lost because words from early in the shadowed message were recognised.

• - These findings support those found by Cherry (1953). 

 - Most participants ignored the instructions that were presented in the passages they were shadowing, and said they thought this was merely an attempt to distract them.

- There should have been 36 sets of instructions preceded by the participant’s name presented in the rejected messsage. However the discrepancy is due to three participants who heard the instructions and actually changed over, so that the second set of instructions which would normally have been heard as part of the shadowed message were now heard as part of the rejected message. These all occurred in passage 10.

-The mean number of instructions heard when presented in the rejected message was calculated, and the difference between the ‘names’ and ‘no names’ was significant: t = 3.05 (significant at greater than the 1 per cent. level, where t = 2.81.

- On only 4 out of the 20 occasions in which the ‘names’ instructions were heard did the participants actually make a change to the other message. 

• The difference between the mean number of digits reported under the two conditions of set were analysed and submitted to a t test. In none of the cases, whether the score was the mean number of digits spoken during shadowing, nor in the number reported, nor the sum of these two was the difference significant even at the 5% level of confidence. 

• In a situation where a participant directs his attention to the reception of a message from one ear, and rejects a message from the other ear, almost none of the verbal content of the rejected message is able to penetrate the block set up.

• A short list of simple words presented as the rejected message shows no trace of being remembered even when presented many times.

• Subjectively ‘important’ messages, such as a person’s own name, can penetrate the block: thus a person will hear instructions if they are presented with their own name as part of the rejected message.

• While perhaps not impossible, it is very difficult to make ‘neutral’ material important enough to break through the block set up in dichotic shadowing.