Practical Research with Children

Jess Prior, Jo Van Herwegen, Jess Prior, Jo Van Herwegen

• Publisher: Taylor & Francis
• Year: 2016
• Print ISBN: 9781138937888
• eBook ISBN: 9781317384045
• Edition: 1
• Topic: Psychology
• Subtopic: Developmental Psychology
• Index: Psychology

Practical Research with Children is designed to help the reader understand techniques for research with children, based on real world experience. The book describes a wide range of research methods, focusing equally on quantitative and qualitative approaches, and considers how different methods can be integrated. It highlights the benefits and challenges of each method and gives emphasis to best practice, with expert guidance on how to avoid potential pitfalls in order to obtain valuable insights into how children develop.

The volume includes fifteen chapters arranged over three sections. Each chapter explores a particular method, or combination of methods, and discusses both theoretical and practical issues, using a diversity of domains, including different ages, cultures, populations and settings. Uniquely, the book includes newer methods (such as eye tracking and digital technologies) alongside well-established behavioural methods which are used for research with children.

With contributions from internationally renowned researchers and practitioners from a range of disciplines, the book will be indispensable reading for a wide audience, including for students in psychology, education and nursing undertaking research projects with children, and also for anyone looking to understand the research behind current theories in child development.

Part 1
Quantitative methods in research with children

1
Researching Cognitive Development in Infancy

Caspar Addyman and Luke Mason

1.1 Introduction

Infancy is the time of greatest change and fastest learning. The brain is still growing and at its most malleable. This is when the foundations of cognition are laid down. Starting from scratch, infants must learn to walk, talk, classify all the objects in their world and learn how to manipulate them. They must learn how to regulate their emotions and interact with other people. Infancy should therefore be of interest to researchers across all areas of psychology. But infants cannot answer questionnaires or press buttons. Their attention span is even worse than undergraduates’. Infants are hard participants to find, to recruit and to schedule. They can be expensive and time consuming to test and only give small amounts of intrinsically noisy data. Research with infants presents considerable challenges and requires substantial ingenuity.

The challenges are both practical and conceptual. How do we run experiments with infants and how do we interpret the results? There are four interrelated challenges that are unique to research with infants. These are the conceptual limitations imposed by infant perception and cognition as well as the practical limitations imposed by infant attention and by infant responding. All four factors may affect an experiment to a greater or lesser degree. Methods in infancy research have been developed largely in response to these constraints. For the youngest ages perceptual limitations are the major consideration and may even be a topic of research in their own right. What is the resolution of infant vision? How well can they localise sounds in space? When do they develop the ability to smoothly track moving objects? Most important from a theoretical point of view are the limitations imposed by infant cognition. Very often these are precisely the questions we are trying to answer. How limited is infant memory? How limited is infant vocabulary, object concepts or category knowledge? How early do abilities arise? How similar or different are they to their adult equivalents? It is crucial to remember that these factors are interrelated. An infant’s very limited working memory could affect their performance in any number of tasks. Object knowledge precedes word learning and so on.

The limitations imposed by infant attention and response options are more practical in nature. The biggest challenge is that infants do not know they are in an experiment. Experiments must therefore be built around infants’ natural behaviours or non-behavioural measures. Behavioural measures are usually visual, namely what infants choose to look at and for how long, but can also involve physical responses like touching or pointing. Non-behavioural measures include physiological markers such as heart rate variation or electrical potentials on the scalp related to underlying brain activity. Responses must be evoked by relatively naturalistic settings that engage infants with simple tasks in a single session of several minutes to half an hour. Whatever the task, infants will get bored very quickly. Indeed, as we discuss below, many classic behavioural paradigms such as habituation exploit this. But infants are not just habituating to the stimuli on the screen, they become bored of the screen itself and of being in the quiet, darkened room where you run the study. Infants’ limited attention makes every experiment a race against time.

The investigation of cognitive development in infancy has also been shaped by the technologies that are available. This chapter therefore takes a historical approach to the field, starting with a survey of classical behavioural methods that capitalise on infants’ tendency to get bored very quickly and to seek out novelty and play games. We go on to describe the revolution brought about by the introduction of eye-trackers and the benefits and difficulties of using electroencephalography (EEG) with infants. We finish by discussing some recent innovations such as Near Infrared Spectroscopy (NIRS).

However, before discussing the experimental methods themselves we take a step back and consider the broader context of research with infants. There are many practical considerations unique to testing with infants and very young children. The first section of this chapter addresses those issues. We consider the designing of a lab space suitable for infant testing, recruiting and getting infants to the lab, making the visit safe and enjoyable for parents and infants.

1.2 Practical considerations

It may be tempting to focus on the methodological and technological requirements of conducting infant research, but the reality is that a variety of other factors begin exerting an influence on infants’ behaviour before they even arrive at the lab.

1.2.1 The physical and social environment

Ensure that the facilities in your lab are appropriate for parents and babies. There should be space to park buggies, toys to play with before the experiment begins (and the ability to disinfect these afterwards), and facilities for baby changing and breastfeeding. In spaces where experimental sessions are filmed, ensure that the cameras can be switched off during feeding.

Infants’ sleep patterns are markedly different from those of children and adults, and vary across age and between individuals. When booking babies for an experiment, it is vital to take naps into account. Parents know their child’s schedule and duration of naps, so gather this information and plan around it. Be aware that an unfamiliar environment, unfamiliar people and the energy-sapping nature of experimental tasks will all work to increase nap duration and decrease the time between naps. A 6-month-old infant who has recently napped has a maximum of 1.5 hours of useful attention available to us as researchers before they need to take a break and, in most cases, to sleep. The clock starts ticking from the moment they walk through the door.

Parents are your greatest allies in gathering good data from their child. Even very young infants are constantly monitoring their parents’ mood and a comfortable and relaxed parent will transmit this state to their child. It is essential to take time to explain clearly to parents what will happen, and why the research their child is taking part in is important. Parents may worry whether their child ‘has passed the test’ or met the requirements of the experiment. Emphasise that researchers are interested in how children naturally respond so there is never a ‘right answer’. Equally, they should not feel pressured to continue if their baby is unhappy. Babies are people too with the same rights as adults and all experiments need to comply with the Helsinki convention and local ethics committees (see Chapter 15 for a detailed discussion of ethical issues when working with children and infants). But it is parents who give their consent and judge that their babies are happy. Unhappy babies give bad data. Remember too, that parental recommendations are a fantastic way of recruiting future participants.

Interacting with the infants themselves requires sensitivity on the part of the researcher and benefits greatly from experience. Just like adults, infants vary in temperament and sociability. Some will approach you immediately to play, others will become very upset if you approach them within 10 minutes of their arrival. Infants and toddlers generally view new people with mistrust until they prove themselves to be friendly, fun and non-threatening; rapport built with parents will usually rub off on infants. Limit the number of people that the infant meets to the bare minimum necessary for running the experiment. Making yourself physically less threatening, by coming down to the level of the child, using infant directed speech, and allowing the infant to come to you are all helpful in warming them up to your presence. An experienced infant researcher once remarked that when interacting with babies you must ‘leave your dignity at the door’!

1.2.2 Maximising attention and compliance

Infants’ attentional capacity is extremely limited and their mood is easily disrupted. Ensure that you set up everything you can in advance of the infant arriving, so that when they move into the lab proper to begin an experiment there is the minimum possible delay. Any technology must be tested and confirmed working before the baby even enters the building – do not let them sit there while a computer boots and cables are plugged in.

It is tempting to play videos to infants before a screen-based task begins, but this is usually a bad idea (although very short videos between tasks do help keep infants engaged). Young infants will habituate to screens and rooms reducing the time they will attend to your experimental stimuli. Older infants and toddlers will enjoy watching their favourite cartoon while an EEG net is applied, but woe betide the researcher who switches off a 2-year-old’s Peppa Pig video and attempts to replace it with boring experimental stimuli! Instead, blow bubbles for younger infants, and use quiet, unstimulating toys with toddlers.

Over the course of the session, infants will rarely remain stationary, often leaning backward, hunching forward, or tilting their upper torso to one side or another (a particular problem for eye-tracking methods; Hessels, Cornelissen, Kemner, & Hooge, 2014). Seating the infant in a high chair or car seat will reduce the range of possible motion compared with having her sat on the parent’s lap, but lap-based babies can be more easily repositioned by their parents during an experiment, and are less likely to fuss out. A middle ground that some researchers adopt is to have the parent wear a front-facing baby carrier, maximising the ability of the parent to intervene while holding the infant in a relatively constrained position in which they feel comfortable and safe.

1.2.3 Design and pacing of stimulus presentation

It is often desirable to use the same underlying paradigm with infants as with adults, in order to acquire comparable data across development. Simply taking an experiment that works with adults and expecting an infant to sit through it, however, is unlikely to bear fruit. Stimuli should be designed in accordance with what is known about a particular age group’s perceptual preferences and abilities. Young infants enjoy watching high contrast, colourful, moving stimuli. Small geometric stimuli on grey backgrounds work well with undergraduates who can be told to sit still and bribed with money or course credit, but an infant will lose interest almost immediately.

Changing the low-level visual features of stimuli such as colour and pattern between blocks, while keeping the underlying structure such as the spatial and temporal arrangement constant, can significantly increase the amount of time an infant will attend to a screen. Trial numbers should be kept to the absolute minimum necessary and rather than presented in one long block they can be interspersed with short, engaging attention-getters, such as a few seconds of pictures or video.

The pacing of stimulus presentation is important, too. Younger infants will take longer to disengage from stimuli and so will be unable to keep up with a fast rate of change. Toddlers, on the other hand, will rapidly grow bored of a slowly changing screen, and will readily direct their attention to running around, pulling on cables or throwing a tantrum.

Many researchers run more than one experiment in any particular session. For example, the learning and recognition stages of a language task could be separated by an unrelated habituation task. This can work effectively, but significantly improves the complexity and number of factors that need to be controlled for. Interspersing blocks of several different tasks can buy an increase in overall attention, but at the expense of not collecting enough data to analyse any of the individual tasks if the infant becomes fussy halfway through the session. Some infants (particularly those that are developing atypically) may strongly dislike certain tasks. This may have a negative effect on other tasks in the same session, so it is strongly advised that you ensure you have the ability to skip or exclude these tasks in such a situation.

Finally, it is vital to pilot all experimental tasks with the target age group. You will make numerous assumptions and balance a variety of constraints when putting together an experimental task, and even very experienced infant researchers are often unable to predict accurately how well a particular task will work. Piloting ensures first that an infant will sit through your task, second that the task produces suitable data from an infant population, and also that you are well practised at running the task so valuable infant participants aren’t lost to experimenter error.

1.3 Classical behavioural methods

Infant research began with Jean Piaget in the 1920s. Piaget took detailed observations of his own children and tried to make them repeatable and interpretable. Typical of this approach was his demonstration of object permanence, whereby infants under 9 months will fail to retrieve a ball hidden under a blanket (Piaget, 1954). The experiment aims to show that for younger infants ‘out of sight is out of mind’. But the task itself does not lend itself to investigating other aspects of infant cognition. Despite his influence on the field, Piaget didn’t introduce systematic methods for understanding infancy.

1.3.1 Looking time methods

A more general approach came about with the introduction of looking time measures by Robert Fantz (Fantz, 1958). The amount of time an infant looks at a stimulus provides a measure of how interesting the infant finds it. This opened up a means of investigating numerous different questions with the same standard methods. Visual preference methods provided a simple means of demonstrating infant discrimination abilities. Habituation allowed for the demonstration and investigation of online learning while violation of expectation probed infants’ beliefs about the world. These methods can also be adapted for use with auditory and cross-modal stimuli. They are represented schematically in Figure 1.1. In visual paradigms it is important to ensure the infant is looking at the screen at the start of each trial. It is usual to use a central flashing object and accompanying sound effect to draw attention back to the screen before each trial begins.

1.3.2 Visual preference

The simplest of these methods is the visual preference test (Fantz, 1958; Teller, 1979). Infants are shown two ite...

Table of contents

• Cover
• Title
• Copyright
• CONTENTS
• List of figures
• List of contributors
• About the editors
• Preface
• Acknowledgements from the editors
• Acknowledgements from the authors
• PART 1 Quantitative methods in research with children
• PART 2 Qualitative methods in research with children
• PART 3 Mixed methods designs in research with children
• Index